diff options
| author | Michael Voronin <survivor.mail@gmail.com> | 2018-05-03 17:12:01 +0300 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2018-05-03 17:12:01 +0300 |
| commit | 5ea967d97a30f0084883d4efa81b05bea3e5d148 (patch) | |
| tree | 05ea0e3624f6720c2f5af28b5a70c87c85feafc7 /tests | |
| parent | 3949c9f977378ea3ab2b3c750f4dc2bc8d853022 (diff) | |
| parent | 5564289b577c620cbd775f477b7fc8b6507adbfa (diff) | |
| download | Nim-5ea967d97a30f0084883d4efa81b05bea3e5d148.tar.gz | |
Merge pull request #3 from nim-lang/devel
pull #3
Diffstat (limited to 'tests')
45 files changed, 347 insertions, 1910 deletions
diff --git a/tests/array/tarraycons_ptr_generic.nim b/tests/array/tarraycons_ptr_generic.nim new file mode 100644 index 000000000..eb89a196f --- /dev/null +++ b/tests/array/tarraycons_ptr_generic.nim @@ -0,0 +1,51 @@ +discard """ + output: '''apple +banana +Fruit +2 +4 +3 +none +skin +paper +''' +""" +type + Fruit = object of RootObj + name: string + Apple = object of Fruit + Banana = object of Fruit + +var + ir = Fruit(name: "Fruit") + ia = Apple(name: "apple") + ib = Banana(name: "banana") + +let x = [ia.addr, ib.addr, ir.addr] +for c in x: echo c.name + +type + Vehicle[T] = object of RootObj + tire: T + Car[T] = object of Vehicle[T] + Bike[T] = object of Vehicle[T] + +var v = Vehicle[int](tire: 3) +var c = Car[int](tire: 4) +var b = Bike[int](tire: 2) + +let y = [b.addr, c.addr, v.addr] +for c in y: echo c.tire + +type + Book[T] = ref object of RootObj + cover: T + Hard[T] = ref object of Book[T] + Soft[T] = ref object of Book[T] + +var bn = Book[string](cover: "none") +var hs = Hard[string](cover: "skin") +var bp = Soft[string](cover: "paper") + +let z = [bn, hs, bp] +for c in z: echo c.cover diff --git a/tests/array/tarraycons_ptr_generic2.nim b/tests/array/tarraycons_ptr_generic2.nim new file mode 100644 index 000000000..fce7af669 --- /dev/null +++ b/tests/array/tarraycons_ptr_generic2.nim @@ -0,0 +1,17 @@ +discard """ + file: "tarraycons_ptr_generic2.nim" + line: 17 + errormsg: "type mismatch: got <ptr Hard[system.string]> but expected 'Book[system.string]'" +""" + +type + Book[T] = ref object of RootObj + cover: T + Hard[T] = ref object of Book[T] + Soft[T] = ref object of Book[T] + +var bn = Book[string](cover: "none") +var hs = Hard[string](cover: "skin") +var bp = Soft[string](cover: "paper") + +let z = [bn, hs.addr, bp] diff --git a/tests/async/tlambda.nim b/tests/async/tlambda.nim index d187c0d50..8f570689b 100644 --- a/tests/async/tlambda.nim +++ b/tests/async/tlambda.nim @@ -1,7 +1,7 @@ # bug 2007 -import asyncdispatch, asyncnet, logging, json, uri, strutils, future +import asyncdispatch, asyncnet, logging, json, uri, strutils, sugar type Builder = ref object @@ -27,7 +27,7 @@ proc newBuild*(onProgress: ProgressCB): Build = result.onProgress = onProgress proc start(build: Build, repo, hash: string) {.async.} = - let path = repo.parseUri().path.toLower() + let path = repo.parseUri().path.toLowerAscii() proc onProgress(builder: Builder, message: string) {.async.} = debug($message) diff --git a/tests/compiles/trecursive_generic_in_compiles.nim b/tests/compiles/trecursive_generic_in_compiles.nim index 77bf0bb02..9c7fd10b3 100644 --- a/tests/compiles/trecursive_generic_in_compiles.nim +++ b/tests/compiles/trecursive_generic_in_compiles.nim @@ -1,6 +1,6 @@ # bug #3313 -import unittest, future - +import unittest, sugar +{.experimental: "notnil".} type ListNodeKind = enum lnkNil, lnkCons diff --git a/tests/concepts/tstackconcept.nim b/tests/concepts/tstackconcept.nim index 2238dacb6..cb8db566d 100644 --- a/tests/concepts/tstackconcept.nim +++ b/tests/concepts/tstackconcept.nim @@ -31,7 +31,7 @@ type s.pop() is T type ValueType = T - const ValueTypeName = T.name.toUpper + const ValueTypeName = T.name.toUpperAscii proc genericAlgorithm[T](s: var Stack[T], y: T) = static: diff --git a/tests/constructors/tinvalid_construction.nim b/tests/constructors/tinvalid_construction.nim index bb3b1bebb..b3e56eec6 100644 --- a/tests/constructors/tinvalid_construction.nim +++ b/tests/constructors/tinvalid_construction.nim @@ -3,12 +3,12 @@ template accept(x) = template reject(x) = static: assert(not compiles(x)) - +{.experimental: "notnil".} type TRefObj = ref object x: int - THasNotNils = object of TObject + THasNotNils = object of RootObj a: TRefObj not nil b: TRefObj not nil c: TRefObj diff --git a/tests/converter/tconvert.nim b/tests/converter/tconvert.nim index a37140234..48367a85b 100644 --- a/tests/converter/tconvert.nim +++ b/tests/converter/tconvert.nim @@ -15,6 +15,6 @@ type TFoo = object converter toPtr*(some: var TFoo): ptr TFoo = (addr some) -proc zoot(x: ptr TFoo) = nil +proc zoot(x: ptr TFoo) = discard var x: Tfoo zoot(x) diff --git a/tests/distinct/tnil.nim b/tests/distinct/tnil.nim index e60437a1f..759a14657 100644 --- a/tests/distinct/tnil.nim +++ b/tests/distinct/tnil.nim @@ -1,15 +1,11 @@ discard """ file: "tnil.nim" - output: '''0x1 - -nil - -nil - + output: '''1 +0 +0 ''' - disabled: "windows" """ - +{.experimental: "notnil".} type MyPointer = distinct pointer MyString = distinct string @@ -17,7 +13,8 @@ type MyInt = distinct int proc foo(a: MyPointer) = - echo a.repr + # workaround a Windows 'repr' difference: + echo cast[int](a) foo(cast[MyPointer](1)) foo(cast[MyPointer](nil)) diff --git a/tests/effects/teffects4.nim b/tests/effects/teffects4.nim index fd5dd49e2..d0960126f 100644 --- a/tests/effects/teffects4.nim +++ b/tests/effects/teffects4.nim @@ -12,7 +12,7 @@ type EIO2 = ref object of EIO proc q() {.tags: [FIO].} = - nil + discard proc raiser(): int = writeLine stdout, "arg" diff --git a/tests/errmsgs/tproper_stacktrace2.nim b/tests/errmsgs/tproper_stacktrace2.nim index 5f312b870..44b208c87 100644 --- a/tests/errmsgs/tproper_stacktrace2.nim +++ b/tests/errmsgs/tproper_stacktrace2.nim @@ -3,7 +3,7 @@ discard """ exitcode: 1 """ -proc returnsNil(): string = return nil +proc returnsNil(): ref int = return nil iterator fields*(a, b: int): int = if a == b: @@ -17,6 +17,6 @@ proc main(): string = result = "" for i in fields(0, 1): let x = returnsNil() - result &= "string literal " & $x + result &= "string literal " & $x[] echo main() diff --git a/tests/manyloc/argument_parser/argument_parser.nim b/tests/manyloc/argument_parser/argument_parser.nim index 14352066d..1095a893e 100644 --- a/tests/manyloc/argument_parser/argument_parser.nim +++ b/tests/manyloc/argument_parser/argument_parser.nim @@ -209,7 +209,7 @@ proc `$`*(data: Tcommandline_results): string = # - Parse code -template raise_or_quit(exception, message: expr): stmt {.immediate.} = +template raise_or_quit(exception, message: untyped) = ## Avoids repeating if check based on the default quit_on_failure variable. ## ## As a special case, if message has a zero length the call to quit won't @@ -230,15 +230,15 @@ template run_custom_proc(parsed_parameter: Tparsed_parameter, ## ## Pass in the string of the parameter triggering the call. If the if not custom_validator.isNil: + try: + let message = custom_validator(parameter, parsed_parameter) + if not message.isNil and message.len > 0: + raise_or_quit(ValueError, ("Failed to validate value for " & + "parameter $1:\n$2" % [escape(parameter), message])) except: raise_or_quit(ValueError, ("Couldn't run custom proc for " & "parameter $1:\n$2" % [escape(parameter), getCurrentExceptionMsg()])) - let message = custom_validator(parameter, parsed_parameter) - if not message.isNil and message.len > 0: - raise_or_quit(ValueError, ("Failed to validate value for " & - "parameter $1:\n$2" % [escape(parameter), message])) - proc parse_parameter(quit_on_failure: bool, param, value: string, param_kind: Tparam_kind): Tparsed_parameter = diff --git a/tests/manyloc/keineschweine/dependencies/chipmunk/chipmunk.nim b/tests/manyloc/keineschweine/dependencies/chipmunk/chipmunk.nim index f06c4e0be..ac425c7a0 100644 --- a/tests/manyloc/keineschweine/dependencies/chipmunk/chipmunk.nim +++ b/tests/manyloc/keineschweine/dependencies/chipmunk/chipmunk.nim @@ -388,13 +388,13 @@ type cdecl.} ##cp property emulators -template defGetter(otype: typedesc, memberType: typedesc, memberName: expr, procName: expr): stmt {.immediate.} = +template defGetter(otype: typedesc, memberType: typedesc, memberName, procName: untyped) = proc `get procName`*(obj: otype): memberType {.cdecl.} = return obj.memberName -template defSetter(otype: typedesc, memberType: typedesc, memberName: expr, procName: expr): stmt {.immediate.} = +template defSetter(otype: typedesc, memberType: typedesc, memberName, procName: untyped) = proc `set procName`*(obj: otype, value: memberType) {.cdecl.} = obj.memberName = value -template defProp(otype: typedesc, memberType: typedesc, memberName: expr, procName: expr): stmt {.immediate.} = +template defProp(otype: typedesc, memberType: typedesc, memberName, procName: untyped) = defGetter(otype, memberType, memberName, procName) defSetter(otype, memberType, memberName, procName) @@ -908,7 +908,7 @@ proc getShapes*(arb: PArbiter, a, b: var PShape) {.inline.} = #/ A macro shortcut for defining and retrieving the shapes from an arbiter. #define CP_ARBITER_GET_SHAPES(arb, a, b) cpShape *a, *b; cpArbiterGetShapes(arb, &a, &b); -template getShapes*(arb: PArbiter, name1, name2: expr): stmt {.immediate.} = +template getShapes*(arb: PArbiter, name1, name2: untyped) = var name1, name2: PShape getShapes(arb, name1, name2) @@ -923,7 +923,7 @@ template getShapes*(arb: PArbiter, name1, name2: expr): stmt {.immediate.} = #/ A macro shortcut for defining and retrieving the bodies from an arbiter. #define CP_ARBITER_GET_BODIES(arb, a, b) cpBody *a, *b; cpArbiterGetBodies(arb, &a, &b); -template getBodies*(arb: PArbiter, name1, name2: expr): stmt {.immediate.} = +template getBodies*(arb: PArbiter, name1, name2: untyped) = var name1, name2: PBOdy getBodies(arb, name1, name2) @@ -947,11 +947,11 @@ proc getDepth*(arb: PArbiter; i: cint): CpFloat {. cdecl, importc: "cpArbiterGetDepth", dynlib: Lib.} ##Shapes -template defShapeSetter(memberType: typedesc, memberName: expr, procName: expr, activates: bool): stmt {.immediate.} = +template defShapeSetter(memberType: typedesc, memberName: untyped, procName: untyped, activates: bool) = proc `set procName`*(obj: PShape, value: memberType) {.cdecl.} = if activates and obj.body != nil: obj.body.activate() obj.memberName = value -template defShapeProp(memberType: typedesc, memberName: expr, procName: expr, activates: bool): stmt {.immediate.} = +template defShapeProp(memberType: typedesc, memberName: untyped, procName: untyped, activates: bool) = defGetter(PShape, memberType, memberName, procName) defShapeSetter(memberType, memberName, procName, activates) @@ -1272,11 +1272,11 @@ proc activateBodies(constraint: PConstraint) {.inline.} = # cpConstraintActivateBodies(constraint); \ # constraint->member = value; \ # } -template defConstraintSetter(memberType: typedesc, member: expr, name: expr): stmt {.immediate.} = +template defConstraintSetter(memberType: typedesc, member, name: untyped) = proc `set name`*(constraint: PConstraint, value: memberType) {.cdecl.} = activateBodies(constraint) constraint.member = value -template defConstraintProp(memberType: typedesc, member: expr, name: expr): stmt {.immediate.} = +template defConstraintProp(memberType: typedesc, member, name: untyped) = defGetter(PConstraint, memberType, member, name) defConstraintSetter(memberType, member, name) # CP_DefineConstraintStructGetter(cpSpace*, CP_PRIVATE(space), Space) @@ -1306,18 +1306,18 @@ proc getImpulse*(constraint: PConstraint): CpFloat {.inline.} = # cpConstraintActivateBodies(constraint); \ # ((struct *)constraint)->member = value; \ # } -template constraintCheckCast(constraint: PConstraint, ctype: expr): stmt {.immediate.} = +template constraintCheckCast(constraint: PConstraint, ctype: untyped) = assert(constraint.klass == `ctype getClass`(), "Constraint is the wrong class") -template defCGetter(ctype: expr, memberType: typedesc, member: expr, name: expr): stmt {.immediate.} = +template defCGetter(ctype: untyped, memberType: typedesc, member, name: untyped) = proc `get ctype name`*(constraint: PConstraint): memberType {.cdecl.} = constraintCheckCast(constraint, ctype) result = cast[`P ctype`](constraint).member -template defCSetter(ctype: expr, memberType: typedesc, member: expr, name: expr): stmt {.immediate.} = +template defCSetter(ctype: untyped, memberType: typedesc, member, name: untyped) = proc `set ctype name`*(constraint: PConstraint, value: memberType) {.cdecl.} = constraintCheckCast(constraint, ctype) activateBodies(constraint) cast[`P ctype`](constraint).member = value -template defCProp(ctype: expr, memberType: typedesc, member: expr, name: expr): stmt {.immediate.} = +template defCProp(ctype: untyped, memberType: typedesc, member, name: untyped) = defCGetter(ctype, memberType, member, name) defCSetter(ctype, memberType, member, name) diff --git a/tests/manyloc/keineschweine/dependencies/enet/enet.nim b/tests/manyloc/keineschweine/dependencies/enet/enet.nim index 3ea8172d5..07079f2ea 100644 --- a/tests/manyloc/keineschweine/dependencies/enet/enet.nim +++ b/tests/manyloc/keineschweine/dependencies/enet/enet.nim @@ -24,7 +24,7 @@ const ENET_VERSION_MAJOR* = 1 ENET_VERSION_MINOR* = 3 ENET_VERSION_PATCH* = 3 -template ENET_VERSION_CREATE(major, minor, patch: expr): expr = +template ENET_VERSION_CREATE(major, minor, patch: untyped): untyped = (((major) shl 16) or ((minor) shl 8) or (patch)) const @@ -277,22 +277,22 @@ when defined(Linux) or true: dataLength*: csize TENetSocketSet* = Tfd_set ## see if these are different on win32, if not then get rid of these - template ENET_HOST_TO_NET_16*(value: expr): expr = + template ENET_HOST_TO_NET_16*(value: untyped): untyped = (htons(value)) - template ENET_HOST_TO_NET_32*(value: expr): expr = + template ENET_HOST_TO_NET_32*(value: untyped): untyped = (htonl(value)) - template ENET_NET_TO_HOST_16*(value: expr): expr = + template ENET_NET_TO_HOST_16*(value: untyped): untyped = (ntohs(value)) - template ENET_NET_TO_HOST_32*(value: expr): expr = + template ENET_NET_TO_HOST_32*(value: untyped): untyped = (ntohl(value)) - template ENET_SOCKETSET_EMPTY*(sockset: expr): expr = + template ENET_SOCKETSET_EMPTY*(sockset: untyped): untyped = FD_ZERO(addr((sockset))) - template ENET_SOCKETSET_ADD*(sockset, socket: expr): expr = + template ENET_SOCKETSET_ADD*(sockset, socket: untyped): untyped = FD_SET(socket, addr((sockset))) - template ENET_SOCKETSET_REMOVE*(sockset, socket: expr): expr = + template ENET_SOCKETSET_REMOVE*(sockset, socket: untyped): untyped = FD_CLEAR(socket, addr((sockset))) - template ENET_SOCKETSET_CHECK*(sockset, socket: expr): expr = + template ENET_SOCKETSET_CHECK*(sockset, socket: untyped): untyped = FD_ISSET(socket, addr((sockset))) when defined(Windows): @@ -606,7 +606,7 @@ proc protocolCommandSize*(commandNumber: cuchar): csize{. {.pop.} -from hashes import `!$`, `!&`, THash, hash -proc hash*(x: TAddress): THash {.nimcall, noSideEffect.} = +from hashes import `!$`, `!&`, Hash, hash +proc hash*(x: TAddress): Hash {.nimcall, noSideEffect.} = result = !$(hash(x.host.int32) !& hash(x.port.int16)) diff --git a/tests/manyloc/keineschweine/dependencies/genpacket/genpacket_enet.nim b/tests/manyloc/keineschweine/dependencies/genpacket/genpacket_enet.nim index 142b190ab..3fb4dc7d9 100644 --- a/tests/manyloc/keineschweine/dependencies/genpacket/genpacket_enet.nim +++ b/tests/manyloc/keineschweine/dependencies/genpacket/genpacket_enet.nim @@ -1,15 +1,15 @@ import macros, macro_dsl, estreams from strutils import format -template newLenName(): stmt {.immediate.} = +template newLenName() = let lenName {.inject.} = ^("len"& $lenNames) inc(lenNames) -template defPacketImports*(): stmt {.immediate, dirty.} = +template defPacketImports*() {.dirty.} = import macros, macro_dsl, estreams from strutils import format -macro defPacket*(typeNameN: expr, typeFields: expr): stmt {.immediate.} = +macro defPacket*(typeNameN: untyped, typeFields: untyped): untyped = result = newNimNode(nnkStmtList) let typeName = quoted2ident(typeNameN) @@ -60,7 +60,7 @@ macro defPacket*(typeNameN: expr, typeFields: expr): stmt {.immediate.} = let name = typeFields[i][0] dotName = packetID.dot(name) - resName = newIdentNode(!"result").dot(name) + resName = newIdentNode("result").dot(name) case typeFields[i][1].kind of nnkBracketExpr: #ex: paddedstring[32, '\0'], array[range, type] case $typeFields[i][1][0].ident @@ -141,7 +141,7 @@ macro defPacket*(typeNameN: expr, typeFields: expr): stmt {.immediate.} = const emptyFields = {nnkEmpty, nnkNilLit} var objFields = newNimNode(nnkRecList) - for i in 0.. < len(typeFields): + for i in 0 ..< len(typeFields): let fname = typeFields[i][0] constructorParams.add(newNimNode(nnkIdentDefs).und( fname, @@ -200,7 +200,7 @@ proc iddefs*(a: string; b: NimNode): NimNode {.compileTime.} = proc varTy*(a: NimNode): NimNode {.compileTime.} = result = newNimNode(nnkVarTy).und(a) -macro forwardPacket*(typeName: expr, underlyingType: expr): stmt {.immediate.} = +macro forwardPacket*(typeName: untyped, underlyingType: untyped): untyped = var packetID = ^"p" streamID = ^"s" @@ -234,7 +234,7 @@ macro forwardPacket*(typeName: expr, underlyingType: expr): stmt {.immediate.} = echo "unknown type:", repr(underlyingtype) echo(repr(result)) -template forwardPacketT*(typeName: expr; underlyingType: expr): stmt {.dirty, immediate.} = +template forwardPacketT*(typeName: untyped; underlyingType: untyped) {.dirty.} = proc `read typeName`*(buffer: PBuffer): typeName = #discard readData(s, addr result, sizeof(result)) var res: underlyingType diff --git a/tests/manyloc/keineschweine/dependencies/genpacket/macro_dsl.nim b/tests/manyloc/keineschweine/dependencies/genpacket/macro_dsl.nim index 86c12fbb0..33d2a7177 100644 --- a/tests/manyloc/keineschweine/dependencies/genpacket/macro_dsl.nim +++ b/tests/manyloc/keineschweine/dependencies/genpacket/macro_dsl.nim @@ -10,7 +10,7 @@ proc und*(a: NimNode; b: varargs[NimNode]): NimNode {.compileTime.} = proc `^`*(a: string): NimNode {.compileTime.} = ## new ident node - result = newIdentNode(!a) + result = newIdentNode(a) proc `[]`*(a, b: NimNode): NimNode {.compileTime.} = ## new bracket expression: node[node] not to be confused with node[indx] result = newNimNode(nnkBracketExpr).und(a, b) @@ -34,7 +34,7 @@ proc emptyNode*(): NimNode {.compileTime.} = proc dot*(left, right: NimNode): NimNode {.compileTime.} = result = newNimNode(nnkDotExpr).und(left, right) proc prefix*(a: string, b: NimNode): NimNode {.compileTime.} = - result = newNimNode(nnkPrefix).und(newIdentNode(!a), b) + result = newNimNode(nnkPrefix).und(newIdentNode(a), b) proc quoted2ident*(a: NimNode): NimNode {.compileTime.} = if a.kind != nnkAccQuoted: @@ -45,13 +45,13 @@ proc quoted2ident*(a: NimNode): NimNode {.compileTime.} = result = ^pname -macro `?`(a: expr): expr = +macro `?`(a: untyped): untyped = ## Character literal ?A #=> 'A' result = ($a[1].ident)[0].lit ## echo(?F,?a,?t,?t,?y) when isMainModule: - macro foo(x: stmt): stmt = + macro foo(x: untyped) = result = newNimNode(nnkStmtList) result.add(newNimNode(nnkCall).und(!!"echo", "Hello thar".lit)) result.add(newCall("echo", lit("3 * 45 = "), (3.lit.infix("*", 45.lit)))) diff --git a/tests/manyloc/keineschweine/dependencies/nake/nakefile.nim b/tests/manyloc/keineschweine/dependencies/nake/nakefile.nim index bdf2139c9..5490211de 100644 --- a/tests/manyloc/keineschweine/dependencies/nake/nakefile.nim +++ b/tests/manyloc/keineschweine/dependencies/nake/nakefile.nim @@ -7,7 +7,7 @@ task "install", "compile and install nake binary": for index, dir in pairs(path): echo " ", index, ". ", dir echo "Where to install nake binary? (quit with ^C or quit or exit)" - let ans = stdin.readLine().toLower + let ans = stdin.readLine().toLowerAscii var index = 0 case ans of "q", "quit", "x", "exit": diff --git a/tests/manyloc/keineschweine/lib/estreams.nim b/tests/manyloc/keineschweine/lib/estreams.nim index bdf9b2bf0..00a55c626 100644 --- a/tests/manyloc/keineschweine/lib/estreams.nim +++ b/tests/manyloc/keineschweine/lib/estreams.nim @@ -78,7 +78,7 @@ proc write*(buffer: PBuffer; val: var string) = setLen buffer.data, buffer.pos + length.int copyMem(addr buffer.data[buffer.pos], addr val[0], length.int) inc buffer.pos, length.int -proc write*[T: TNumber|bool|char|byte](buffer: PBuffer; val: T) = +proc write*[T: SomeNumber|bool|char|byte](buffer: PBuffer; val: T) = var v: T shallowCopy v, val writeBE buffer, v diff --git a/tests/manyloc/keineschweine/lib/input_helpers.nim b/tests/manyloc/keineschweine/lib/input_helpers.nim index 1953cb58c..ff1286c8d 100644 --- a/tests/manyloc/keineschweine/lib/input_helpers.nim +++ b/tests/manyloc/keineschweine/lib/input_helpers.nim @@ -5,8 +5,8 @@ type TInputFinishedProc* = proc() TKeyCallback = proc() PKeyClient* = ref object - onKeyDown: TTable[int32, TKeyCallback] - onKeyUp: TTable[int32, TKeyCallback] + onKeyDown: Table[int32, TKeyCallback] + onKeyUp: Table[int32, TKeyCallback] name: string PTextInput* = ref object text*: string @@ -134,5 +134,5 @@ iterator pollEvents*(window: PRenderWindow): PEvent = of EvtMouseButtonReleased: addButtonEvent(event.mouseButton.button, up) of EvtTextEntered: recordText(activeInput, event.text) of EvtMouseMoved: setMousePos(event.mouseMove.x, event.mouseMove.y) - else: nil + else: discard yield(addr event) diff --git a/tests/manyloc/keineschweine/lib/sg_assets.nim b/tests/manyloc/keineschweine/lib/sg_assets.nim index fbc3c9ab8..1e8a99c83 100644 --- a/tests/manyloc/keineschweine/lib/sg_assets.nim +++ b/tests/manyloc/keineschweine/lib/sg_assets.nim @@ -59,7 +59,7 @@ type of Projectile: bullet*: PBulletRecord else: - nil + discard PBulletRecord* = ref TBulletRecord TBulletRecord* = object id*: int16 @@ -115,10 +115,10 @@ var cfg: PZoneSettings SpriteSheets* = initTable[string, PSpriteSheet](64) SoundCache * = initTable[string, PSoundRecord](64) - nameToVehID*: TTable[string, int] - nameToItemID*: TTable[string, int] - nameToObjID*: TTable[string, int] - nameToBulletID*: TTable[string, int] + nameToVehID*: Table[string, int] + nameToItemID*: Table[string, int] + nameToObjID*: Table[string, int] + nameToBulletID*: Table[string, int] activeState = Lobby proc newSprite*(filename: string; errors: var seq[string]): PSpriteSheet @@ -126,7 +126,7 @@ proc load*(ss: PSpriteSheet): bool {.discardable.} proc newSound*(filename: string; errors: var seq[string]): PSoundRecord proc load*(s: PSoundRecord): bool {.discardable.} -proc validateSettings*(settings: PJsonNode; errors: var seq[string]): bool +proc validateSettings*(settings: JsonNode; errors: var seq[string]): bool proc loadSettings*(rawJson: string, errors: var seq[string]): bool proc loadSettingsFromFile*(filename: string, errors: var seq[string]): bool @@ -135,17 +135,17 @@ proc fetchItm*(itm: string): PItemRecord proc fetchObj*(name: string): PObjectRecord proc fetchBullet(name: string): PBulletRecord -proc importLevel(data: PJsonNode; errors: var seq[string]): PLevelSettings -proc importVeh(data: PJsonNode; errors: var seq[string]): PVehicleRecord -proc importObject(data: PJsonNode; errors: var seq[string]): PObjectRecord -proc importItem(data: PJsonNode; errors: var seq[string]): PItemRecord -proc importPhys(data: PJsonNode): TPhysicsRecord -proc importAnim(data: PJsonNode; errors: var seq[string]): PAnimationRecord -proc importHandling(data: PJsonNode): THandlingRecord -proc importBullet(data: PJsonNode; errors: var seq[string]): PBulletRecord -proc importSoul(data: PJsonNode): TSoulRecord -proc importExplosion(data: PJsonNode; errors: var seq[string]): TExplosionRecord -proc importSound*(data: PJsonNode; errors: var seq[string]; fieldName: string = nil): PSoundRecord +proc importLevel(data: JsonNode; errors: var seq[string]): PLevelSettings +proc importVeh(data: JsonNode; errors: var seq[string]): PVehicleRecord +proc importObject(data: JsonNode; errors: var seq[string]): PObjectRecord +proc importItem(data: JsonNode; errors: var seq[string]): PItemRecord +proc importPhys(data: JsonNode): TPhysicsRecord +proc importAnim(data: JsonNode; errors: var seq[string]): PAnimationRecord +proc importHandling(data: JsonNode): THandlingRecord +proc importBullet(data: JsonNode; errors: var seq[string]): PBulletRecord +proc importSoul(data: JsonNode): TSoulRecord +proc importExplosion(data: JsonNode; errors: var seq[string]): TExplosionRecord +proc importSound*(data: JsonNode; errors: var seq[string]; fieldName: string = nil): PSoundRecord ## this is the only pipe between lobby and main.nim proc getActiveState*(): TGameState = @@ -203,7 +203,7 @@ iterator playableVehicles*(): PVehicleRecord = if v.playable: yield v -template allAssets*(body: stmt) {.dirty.}= +template allAssets*(body: untyped) {.dirty.}= block: var assetType = FGraphics for file, asset in pairs(SpriteSheets): @@ -212,7 +212,7 @@ template allAssets*(body: stmt) {.dirty.}= for file, asset in pairs(SoundCache): body -template cacheImpl(procName, cacheName, resultType: expr; body: stmt) {.dirty, immediate.} = +template cacheImpl(procName, cacheName, resultType, body: untyped) {.dirty.} = proc procName*(filename: string; errors: var seq[string]): resulttype = if hasKey(cacheName, filename): return cacheName[filename] @@ -220,7 +220,7 @@ template cacheImpl(procName, cacheName, resultType: expr; body: stmt) {.dirty, i body cacheName[filename] = result -template checkFile(path: expr): stmt {.dirty, immediate.} = +template checkFile(path: untyped) {.dirty.} = if not existsFile(path): errors.add("File missing: "& path) @@ -243,7 +243,7 @@ proc expandPath*(assetType: TAssetType; fileName: string): string = case assetType of FGraphics: result.add "gfx/" of FSound: result.add "sfx/" - else: nil + else: discard result.add fileName proc expandPath*(fc: ScFileChallenge): string {.inline.} = result = expandPath(fc.assetType, fc.file) @@ -280,10 +280,10 @@ else: if not s.soundBuf.isNil: result = true -template addError(e: expr): stmt {.immediate.} = +template addError(e: untyped) = errors.add(e) result = false -proc validateSettings*(settings: PJsonNode, errors: var seq[string]): bool = +proc validateSettings*(settings: JsonNode, errors: var seq[string]): bool = result = true if settings.kind != JObject: addError("Settings root must be an object") @@ -328,10 +328,10 @@ proc loadSettingsFromFile*(filename: string, errors: var seq[string]): bool = result = loadSettings(readFile(filename), errors) proc loadSettings*(rawJson: string, errors: var seq[string]): bool = - var settings: PJsonNode + var settings: JsonNode try: settings = parseJson(rawJson) - except EJsonParsingError: + except JsonParsingError: errors.add("JSON parsing error: "& getCurrentExceptionMsg()) return except: @@ -407,21 +407,21 @@ proc fetchObj*(name: string): PObjectRecord = proc fetchBullet(name: string): PBulletRecord = return cfg.bullets[nameToBulletID[name]] -proc getField(node: PJsonNode, field: string, target: var float) = +proc getField(node: JsonNode, field: string, target: var float) = if not node.hasKey(field): return if node[field].kind == JFloat: target = node[field].fnum elif node[field].kind == JInt: target = node[field].num.float -proc getField(node: PJsonNode, field: string, target: var int) = +proc getField(node: JsonNode, field: string, target: var int) = if not node.hasKey(field): return if node[field].kind == JInt: target = node[field].num.int elif node[field].kind == JFloat: target = node[field].fnum.int -proc getField(node: PJsonNode; field: string; target: var bool) = +proc getField(node: JsonNode; field: string; target: var bool) = if not node.hasKey(field): return case node[field].kind @@ -431,19 +431,19 @@ proc getField(node: PJsonNode; field: string; target: var bool) = target = (node[field].num != 0) of JFloat: target = (node[field].fnum != 0.0) - else: nil + else: discard -template checkKey(node: expr; key: string): stmt = +template checkKey(node: untyped; key: string) = if not hasKey(node, key): return -proc importTrail(data: PJsonNode; errors: var seq[string]): TTrailRecord = +proc importTrail(data: JsonNode; errors: var seq[string]): TTrailRecord = checkKey(data, "trail") result.anim = importAnim(data["trail"], errors) result.timer = 1000.0 getField(data["trail"], "timer", result.timer) result.timer /= 1000.0 -proc importLevel(data: PJsonNode; errors: var seq[string]): PLevelSettings = +proc importLevel(data: JsonNode; errors: var seq[string]): PLevelSettings = new(result) result.size = vec2i(5000, 5000) result.starfield = @[] @@ -456,7 +456,7 @@ proc importLevel(data: PJsonNode; errors: var seq[string]): PLevelSettings = if level.hasKey("starfield"): for star in level["starfield"].items: result.starfield.add(newSprite(star.str, errors)) -proc importPhys(data: PJsonNode): TPhysicsRecord = +proc importPhys(data: JsonNode): TPhysicsRecord = result.radius = 20.0 result.mass = 10.0 @@ -466,7 +466,7 @@ proc importPhys(data: PJsonNode): TPhysicsRecord = phys.getField("mass", result.mass) when not defined(NoChipmunk): result.moment = momentForCircle(result.mass, 0.0, result.radius, VectorZero) * MomentMult -proc importHandling(data: PJsonNode): THandlingRecord = +proc importHandling(data: JsonNode): THandlingRecord = result.thrust = 45.0 result.topSpeed = 100.0 #unused result.reverse = 30.0 @@ -483,7 +483,7 @@ proc importHandling(data: PJsonNode): THandlingRecord = hand.getField("reverse", result.reverse) hand.getField("strafe", result.strafe) hand.getField("rotation", result.rotation) -proc importAnim(data: PJsonNode, errors: var seq[string]): PAnimationRecord = +proc importAnim(data: JsonNode, errors: var seq[string]): PAnimationRecord = new(result) result.angle = 0.0 result.delay = 1000.0 @@ -502,26 +502,26 @@ proc importAnim(data: PJsonNode, errors: var seq[string]): PAnimationRecord = result.angle = radians(result.angle) ## comes in as degrees result.delay /= 1000 ## delay comes in as milliseconds -proc importSoul(data: PJsonNode): TSoulRecord = +proc importSoul(data: JsonNode): TSoulRecord = result.energy = 10000 result.health = 1 checkKey(data, "soul") let soul = data["soul"] soul.getField("energy", result.energy) soul.getField("health", result.health) -proc importExplosion(data: PJsonNode; errors: var seq[string]): TExplosionRecord = +proc importExplosion(data: JsonNode; errors: var seq[string]): TExplosionRecord = checkKey(data, "explode") let expl = data["explode"] result.anim = importAnim(expl, errors) result.sound = importSound(expl, errors, "sound") -proc importSound*(data: PJsonNode; errors: var seq[string]; fieldName: string = nil): PSoundRecord = +proc importSound*(data: JsonNode; errors: var seq[string]; fieldName: string = nil): PSoundRecord = if data.kind == JObject: checkKey(data, fieldName) result = newSound(data[fieldName].str, errors) elif data.kind == JString: result = newSound(data.str, errors) -proc importVeh(data: PJsonNode; errors: var seq[string]): PVehicleRecord = +proc importVeh(data: JsonNode; errors: var seq[string]): PVehicleRecord = new(result) result.playable = false if data.kind != JArray or data.len != 2 or @@ -538,7 +538,7 @@ proc importVeh(data: PJsonNode; errors: var seq[string]): PVehicleRecord = vehdata.getField("playable", result.playable) if result.anim.spriteSheet.isNil and result.playable: result.playable = false -proc importObject(data: PJsonNode; errors: var seq[string]): PObjectRecord = +proc importObject(data: JsonNode; errors: var seq[string]): PObjectRecord = new(result) if data.kind != JArray or data.len != 2: result.name = "(broken)" @@ -546,7 +546,7 @@ proc importObject(data: PJsonNode; errors: var seq[string]): PObjectRecord = result.name = data[0].str result.anim = importAnim(data[1], errors) result.physics = importPhys(data[1]) -proc importItem(data: PJsonNode; errors: var seq[string]): PItemRecord = +proc importItem(data: JsonNode; errors: var seq[string]): PItemRecord = new(result) if data.kind != JArray or data.len != 3: result.name = "(broken)" @@ -562,7 +562,7 @@ proc importItem(data: PJsonNode; errors: var seq[string]): PItemRecord = result.useSound = importSound(data[2], errors, "useSound") - case data[1].str.toLower + case data[1].str.toLowerAscii of "projectile": result.kind = Projectile if data[2]["bullet"].kind == JString: @@ -576,15 +576,15 @@ proc importItem(data: PJsonNode; errors: var seq[string]): PItemRecord = of "ammo": result.kind = Ammo of "utility": - nil + discard else: errors.add "Invalid item type \""&data[1].str&"\" for item "&result.name -proc importBullet(data: PJsonNode; errors: var seq[string]): PBulletRecord = +proc importBullet(data: JsonNode; errors: var seq[string]): PBulletRecord = new(result) result.id = -1 - var bdata: PJsonNode + var bdata: JsonNode if data.kind == JArray: result.name = data[0].str bdata = data[1] diff --git a/tests/manyloc/keineschweine/lib/sg_packets.nim b/tests/manyloc/keineschweine/lib/sg_packets.nim index d84bf72fc..f3a0e8925 100644 --- a/tests/manyloc/keineschweine/lib/sg_packets.nim +++ b/tests/manyloc/keineschweine/lib/sg_packets.nim @@ -4,14 +4,14 @@ defPacketImports() type PacketID* = char -template idpacket(pktName, id, s2c, c2s: expr): stmt {.immediate, dirty.} = +template idpacket(pktName, id, s2c, c2s: untyped) {.dirty.} = let `H pktName`* {.inject.} = id defPacket(`Sc pktName`, s2c) defPacket(`Cs pktName`, c2s) forwardPacketT(uint8, int8) forwardPacketT(uint16, int16) -forwardPacketT(TPort, int16) +forwardPacketT(Port, int16) idPacket(Login, 'a', tuple[id: int32; alias: string; sessionKey: string], @@ -22,7 +22,7 @@ defPacket(CsZoneJoinReq, tuple[session: ScLogin]) defPacket(ScZoneRecord, tuple[ name: string = "", desc: string = "", - ip: string = "", port: TPort = 0.Tport]) + ip: string = "", port: Port = 0.Port]) idPacket(ZoneList, 'z', tuple[network: string = "", zones: seq[ScZoneRecord]], tuple[time: string]) diff --git a/tests/manyloc/nake/nake.nim b/tests/manyloc/nake/nake.nim index 1e88fa73b..728619e47 100644 --- a/tests/manyloc/nake/nake.nim +++ b/tests/manyloc/nake/nake.nim @@ -67,7 +67,7 @@ else: for kind, key, val in getOpt(): case kind of cmdLongOption, cmdShortOption: - case key.tolower + case key.tolowerAscii of "tasks", "t": printTaskList = true else: diff --git a/tests/manyloc/nake/nakefile.nim b/tests/manyloc/nake/nakefile.nim index 97af79a84..3e8609169 100644 --- a/tests/manyloc/nake/nakefile.nim +++ b/tests/manyloc/nake/nakefile.nim @@ -88,7 +88,7 @@ task "download", "download game assets": if existsFile(path): echo "The file already exists\n", "[R]emove [M]ove [Q]uit [S]kip Source: ", GameAssets - case stdin.readLine.toLower + case stdin.readLine.toLowerAscii of "r": removeFile path of "m": @@ -120,7 +120,7 @@ task "download", "download game assets": echo "Download binary libs? Only libs for linux are available currently, enjoy the irony.\n", "[Y]es [N]o Source: ", BinLibs - case stdin.readline.toLower + case stdin.readline.toLowerAscii of "y", "yes": discard ## o_O else: diff --git a/tests/metatype/tbindtypedesc.nim b/tests/metatype/tbindtypedesc.nim index b287aad01..039acfbe9 100644 --- a/tests/metatype/tbindtypedesc.nim +++ b/tests/metatype/tbindtypedesc.nim @@ -46,7 +46,7 @@ type type1 = typedesc type2 = typedesc -proc typePairs(A, B: type1; C, D: type2) = nil +proc typePairs(A, B: type1; C, D: type2) = discard accept typePairs(int, int, TFoo, TFOO) accept typePairs(TBAR, TBar, TBAR, TBAR) @@ -55,7 +55,7 @@ accept typePairs(int, int, string, string) reject typePairs(TBAR, TBar, TBar, TFoo) reject typePairs(string, int, TBAR, TBAR) -proc typePairs2[T: typedesc, U: typedesc](A, B: T; C, D: U) = nil +proc typePairs2[T: typedesc, U: typedesc](A, B: T; C, D: U) = discard accept typePairs2(int, int, TFoo, TFOO) accept typePairs2(TBAR, TBar, TBAR, TBAR) @@ -71,12 +71,12 @@ proc dontBind(a: typedesc, b: typedesc) = accept dontBind(int, float) accept dontBind(TFoo, TFoo) -proc dontBind2(a, b: typedesc) = nil +proc dontBind2(a, b: typedesc) = discard accept dontBind2(int, float) accept dontBind2(TBar, int) -proc bindArg(T: typedesc, U: typedesc, a, b: T, c, d: U) = nil +proc bindArg(T: typedesc, U: typedesc, a, b: T, c, d: U) = discard accept bindArg(int, string, 10, 20, "test", "nest") accept bindArg(int, int, 10, 20, 30, 40) diff --git a/tests/method/tsimmeth.nim b/tests/method/tsimmeth.nim index 11ff2674f..a057c35b7 100644 --- a/tests/method/tsimmeth.nim +++ b/tests/method/tsimmeth.nim @@ -6,7 +6,7 @@ discard """ import strutils -var x = "hello world!".toLower.toUpper +var x = "hello world!".toLowerAscii.toUpperAscii x.echo() #OUT HELLO WORLD! diff --git a/tests/misc/tmemoization.nim b/tests/misc/tmemoization.nim index 180acd89b..840eb3b0d 100644 --- a/tests/misc/tmemoization.nim +++ b/tests/misc/tmemoization.nim @@ -8,7 +8,7 @@ import strutils proc foo(s: static[string]): string = static: echo s - const R = s.toUpper + const R = s.toUpperAscii return R echo foo("test 1") diff --git a/tests/misc/tsemfold.nim b/tests/misc/tsemfold.nim new file mode 100644 index 000000000..18c282d9e --- /dev/null +++ b/tests/misc/tsemfold.nim @@ -0,0 +1,23 @@ +discard """ + action: run +""" + +doAssertRaises(OverflowError): discard low(int8) - 1'i8 +doAssertRaises(OverflowError): discard high(int8) + 1'i8 +doAssertRaises(OverflowError): discard abs(low(int8)) +doAssertRaises(DivByZeroError): discard 1 mod 0 +doAssertRaises(DivByZeroError): discard 1 div 0 +doAssertRaises(OverflowError): discard low(int8) div -1'i8 + +doAssertRaises(OverflowError): discard low(int64) - 1'i64 +doAssertRaises(OverflowError): discard high(int64) + 1'i64 + +type E = enum eA, eB +doAssertRaises(OverflowError): discard eA.pred +doAssertRaises(OverflowError): discard eB.succ + +doAssertRaises(OverflowError): discard low(int8) * -1 +doAssertRaises(OverflowError): discard low(int64) * -1 +doAssertRaises(OverflowError): discard high(int8) * 2 +doAssertRaises(OverflowError): discard high(int64) * 2 + diff --git a/tests/notnil/tmust_compile.nim b/tests/notnil/tmust_compile.nim index 10da154f0..a32c6c7ec 100644 --- a/tests/notnil/tmust_compile.nim +++ b/tests/notnil/tmust_compile.nim @@ -3,6 +3,7 @@ discard """ """ # bug #6682 +{.experimental: "notnil".} type Fields = enum diff --git a/tests/notnil/tnotnil.nim b/tests/notnil/tnotnil.nim index f65634ed6..e392b155c 100644 --- a/tests/notnil/tnotnil.nim +++ b/tests/notnil/tnotnil.nim @@ -2,7 +2,7 @@ discard """ line: 22 errormsg: "type mismatch" """ - +{.experimental: "notnil".} type PObj = ref TObj not nil TObj = object @@ -15,8 +15,8 @@ type proc p(x: string not nil): int = result = 45 -proc q(x: MyString) = nil -proc q2(x: string) = nil +proc q(x: MyString) = discard +proc q2(x: string) = discard q2(nil) q(nil) diff --git a/tests/notnil/tnotnil1.nim b/tests/notnil/tnotnil1.nim index 73472752c..7f9d02295 100644 --- a/tests/notnil/tnotnil1.nim +++ b/tests/notnil/tnotnil1.nim @@ -4,7 +4,7 @@ discard """ """ import strutils - +{.experimental: "notnil".} type TObj = object @@ -18,13 +18,13 @@ proc q(s: superstring) = echo s proc p2() = - var a: string = "I am not nil" + var a: string = "I am not nil" q(a) # but this should and does not p2() proc q(x: pointer not nil) = - nil + discard proc p() = var x: pointer diff --git a/tests/notnil/tnotnil2.nim b/tests/notnil/tnotnil2.nim index bd6b8b675..6cd08de73 100644 --- a/tests/notnil/tnotnil2.nim +++ b/tests/notnil/tnotnil2.nim @@ -4,14 +4,14 @@ discard """ """ import strutils - +{.experimental: "notnil".} type TObj = object x, y: int proc q(x: pointer not nil) = - nil + discard proc p() = var x: pointer diff --git a/tests/notnil/tnotnil3.nim b/tests/notnil/tnotnil3.nim index b7c7a811d..31a4efef7 100644 --- a/tests/notnil/tnotnil3.nim +++ b/tests/notnil/tnotnil3.nim @@ -5,7 +5,7 @@ discard """ # bug #584 # Testprogram for 'not nil' check - +{.experimental: "notnil".} const testWithResult = true type diff --git a/tests/notnil/tnotnil4.nim b/tests/notnil/tnotnil4.nim index 2fa888357..4fd169827 100644 --- a/tests/notnil/tnotnil4.nim +++ b/tests/notnil/tnotnil4.nim @@ -2,6 +2,8 @@ discard "" type TObj = ref object +{.experimental: "notnil".} + proc check(a: TObj not nil) = echo repr(a) diff --git a/tests/notnil/tnotnil_in_generic.nim b/tests/notnil/tnotnil_in_generic.nim index 357ab2c7c..89d20f182 100644 --- a/tests/notnil/tnotnil_in_generic.nim +++ b/tests/notnil/tnotnil_in_generic.nim @@ -3,6 +3,7 @@ discard """ """ # bug #2216 +{.experimental: "notnil".} type A[T] = ref object diff --git a/tests/notnil/tnotnil_in_objconstr.nim b/tests/notnil/tnotnil_in_objconstr.nim index 7dce98c29..d33709906 100644 --- a/tests/notnil/tnotnil_in_objconstr.nim +++ b/tests/notnil/tnotnil_in_objconstr.nim @@ -2,7 +2,7 @@ discard """ errormsg: "fields not initialized: bar" line: "13" """ - +{.experimental: "notnil".} # bug #2355 type Foo = object diff --git a/tests/objvariant/tcheckedfield1.nim b/tests/objvariant/tcheckedfield1.nim index 56d784a2b..a7f232c5b 100644 --- a/tests/objvariant/tcheckedfield1.nim +++ b/tests/objvariant/tcheckedfield1.nim @@ -6,7 +6,7 @@ discard """ import strutils {.warning[ProveField]: on.} - +{.experimental: "notnil".} type TNodeKind = enum nkBinary, nkTernary, nkStr diff --git a/tests/overload/tissue966.nim b/tests/overload/tissue966.nim index c5b28e217..d0a723875 100644 --- a/tests/overload/tissue966.nim +++ b/tests/overload/tissue966.nim @@ -5,7 +5,7 @@ discard """ type PTest = ref object -proc test(x: PTest, y: int) = nil +proc test(x: PTest, y: int) = discard var buf: PTest buf.test() diff --git a/tests/proc/tprocredef.nim b/tests/proc/tprocredef.nim index 86ed92b62..4ec771510 100644 --- a/tests/proc/tprocredef.nim +++ b/tests/proc/tprocredef.nim @@ -4,6 +4,6 @@ discard """ errormsg: "redefinition of \'foo\'" """ -proc foo(a: int, b: string) = nil -proc foo(a: int, b: string) = nil +proc foo(a: int, b: string) = discard +proc foo(a: int, b: string) = discard diff --git a/tests/range/tsubrange.nim b/tests/range/tsubrange.nim index 6f88c5a22..914e7c6e7 100644 --- a/tests/range/tsubrange.nim +++ b/tests/range/tsubrange.nim @@ -7,7 +7,7 @@ type TRange = range[0..40] proc p(r: TRange) = - nil + discard var r: TRange diff --git a/tests/stdlib/tpegs.nim b/tests/stdlib/tpegs.nim deleted file mode 100644 index e2a5a1715..000000000 --- a/tests/stdlib/tpegs.nim +++ /dev/null @@ -1,1770 +0,0 @@ -discard """ - output: '''this -is -an -example -d -e -f -('keyvalue' 'key'*)''' -""" -# PEGs module turned out to be a good test to detect memory management bugs. - -include "system/inclrtl" - -const - useUnicode = true ## change this to deactivate proper UTF-8 support - -import - strutils - -when useUnicode: - import unicode - -const - InlineThreshold = 5 ## number of leaves; -1 to disable inlining - MaxSubpatterns* = 10 ## defines the maximum number of subpatterns that - ## can be captured. More subpatterns cannot be captured! - -type - TPegKind = enum - pkEmpty, - pkAny, ## any character (.) - pkAnyRune, ## any Unicode character (_) - pkNewLine, ## CR-LF, LF, CR - pkLetter, ## Unicode letter - pkLower, ## Unicode lower case letter - pkUpper, ## Unicode upper case letter - pkTitle, ## Unicode title character - pkWhitespace, ## Unicode whitespace character - pkTerminal, - pkTerminalIgnoreCase, - pkTerminalIgnoreStyle, - pkChar, ## single character to match - pkCharChoice, - pkNonTerminal, - pkSequence, ## a b c ... --> Internal DSL: peg(a, b, c) - pkOrderedChoice, ## a / b / ... --> Internal DSL: a / b or /[a, b, c] - pkGreedyRep, ## a* --> Internal DSL: *a - ## a+ --> (a a*) - pkGreedyRepChar, ## x* where x is a single character (superop) - pkGreedyRepSet, ## [set]* (superop) - pkGreedyAny, ## .* or _* (superop) - pkOption, ## a? --> Internal DSL: ?a - pkAndPredicate, ## &a --> Internal DSL: &a - pkNotPredicate, ## !a --> Internal DSL: !a - pkCapture, ## {a} --> Internal DSL: capture(a) - pkBackRef, ## $i --> Internal DSL: backref(i) - pkBackRefIgnoreCase, - pkBackRefIgnoreStyle, - pkSearch, ## @a --> Internal DSL: !*a - pkCapturedSearch, ## {@} a --> Internal DSL: !*\a - pkRule, ## a <- b - pkList, ## a, b - pkStartAnchor ## ^ --> Internal DSL: startAnchor() - TNonTerminalFlag = enum - ntDeclared, ntUsed - TNonTerminal {.final.} = object ## represents a non terminal symbol - name: string ## the name of the symbol - line: int ## line the symbol has been declared/used in - col: int ## column the symbol has been declared/used in - flags: set[TNonTerminalFlag] ## the nonterminal's flags - rule: TNode ## the rule that the symbol refers to - TNode {.final, shallow.} = object - case kind: TPegKind - of pkEmpty..pkWhitespace: nil - of pkTerminal, pkTerminalIgnoreCase, pkTerminalIgnoreStyle: term: string - of pkChar, pkGreedyRepChar: ch: char - of pkCharChoice, pkGreedyRepSet: charChoice: ref set[char] - of pkNonTerminal: nt: PNonTerminal - of pkBackRef..pkBackRefIgnoreStyle: index: range[0..MaxSubpatterns] - else: sons: seq[TNode] - PNonTerminal* = ref TNonTerminal - - TPeg* = TNode ## type that represents a PEG - -proc term*(t: string): TPeg {.rtl, extern: "npegs$1Str".} = - ## constructs a PEG from a terminal string - if t.len != 1: - result.kind = pkTerminal - result.term = t - else: - result.kind = pkChar - result.ch = t[0] - -proc termIgnoreCase*(t: string): TPeg {. - rtl, extern: "npegs$1".} = - ## constructs a PEG from a terminal string; ignore case for matching - result.kind = pkTerminalIgnoreCase - result.term = t - -proc termIgnoreStyle*(t: string): TPeg {. - rtl, extern: "npegs$1".} = - ## constructs a PEG from a terminal string; ignore style for matching - result.kind = pkTerminalIgnoreStyle - result.term = t - -proc term*(t: char): TPeg {.rtl, extern: "npegs$1Char".} = - ## constructs a PEG from a terminal char - assert t != '\0' - result.kind = pkChar - result.ch = t - -proc charSet*(s: set[char]): TPeg {.rtl, extern: "npegs$1".} = - ## constructs a PEG from a character set `s` - assert '\0' notin s - result.kind = pkCharChoice - new(result.charChoice) - result.charChoice[] = s - -proc len(a: TPeg): int {.inline.} = return a.sons.len -proc add(d: var TPeg, s: TPeg) {.inline.} = add(d.sons, s) - -proc copyPeg(a: TPeg): TPeg = - result.kind = a.kind - case a.kind - of pkEmpty..pkWhitespace: discard - of pkTerminal, pkTerminalIgnoreCase, pkTerminalIgnoreStyle: - result.term = a.term - of pkChar, pkGreedyRepChar: - result.ch = a.ch - of pkCharChoice, pkGreedyRepSet: - new(result.charChoice) - result.charChoice[] = a.charChoice[] - of pkNonTerminal: result.nt = a.nt - of pkBackRef..pkBackRefIgnoreStyle: - result.index = a.index - else: - result.sons = a.sons - -proc addChoice(dest: var TPeg, elem: TPeg) = - var L = dest.len-1 - if L >= 0 and dest.sons[L].kind == pkCharChoice: - # caution! Do not introduce false aliasing here! - case elem.kind - of pkCharChoice: - dest.sons[L] = charSet(dest.sons[L].charChoice[] + elem.charChoice[]) - of pkChar: - dest.sons[L] = charSet(dest.sons[L].charChoice[] + {elem.ch}) - else: add(dest, elem) - else: add(dest, elem) - -template multipleOp(k: TPegKind, localOpt) = - result.kind = k - result.sons = @[] - for x in items(a): - if x.kind == k: - for y in items(x.sons): - localOpt(result, y) - else: - localOpt(result, x) - if result.len == 1: - result = result.sons[0] - -proc `/`*(a: varargs[TPeg]): TPeg {. - rtl, extern: "npegsOrderedChoice".} = - ## constructs an ordered choice with the PEGs in `a` - multipleOp(pkOrderedChoice, addChoice) - -proc addSequence(dest: var TPeg, elem: TPeg) = - var L = dest.len-1 - if L >= 0 and dest.sons[L].kind == pkTerminal: - # caution! Do not introduce false aliasing here! - case elem.kind - of pkTerminal: - dest.sons[L] = term(dest.sons[L].term & elem.term) - of pkChar: - dest.sons[L] = term(dest.sons[L].term & elem.ch) - else: add(dest, elem) - else: add(dest, elem) - -proc sequence*(a: varargs[TPeg]): TPeg {. - rtl, extern: "npegs$1".} = - ## constructs a sequence with all the PEGs from `a` - multipleOp(pkSequence, addSequence) - -proc `?`*(a: TPeg): TPeg {.rtl, extern: "npegsOptional".} = - ## constructs an optional for the PEG `a` - if a.kind in {pkOption, pkGreedyRep, pkGreedyAny, pkGreedyRepChar, - pkGreedyRepSet}: - # a* ? --> a* - # a? ? --> a? - result = a - else: - result.kind = pkOption - result.sons = @[a] - -proc `*`*(a: TPeg): TPeg {.rtl, extern: "npegsGreedyRep".} = - ## constructs a "greedy repetition" for the PEG `a` - case a.kind - of pkGreedyRep, pkGreedyRepChar, pkGreedyRepSet, pkGreedyAny, pkOption: - assert false - # produces endless loop! - of pkChar: - result.kind = pkGreedyRepChar - result.ch = a.ch - of pkCharChoice: - result.kind = pkGreedyRepSet - result.charChoice = a.charChoice # copying a reference suffices! - of pkAny, pkAnyRune: - result.kind = pkGreedyAny - else: - result.kind = pkGreedyRep - result.sons = @[a] - -proc `!*`*(a: TPeg): TPeg {.rtl, extern: "npegsSearch".} = - ## constructs a "search" for the PEG `a` - result.kind = pkSearch - result.sons = @[a] - -proc `!*\`*(a: TPeg): TPeg {.rtl, - extern: "npgegsCapturedSearch".} = - ## constructs a "captured search" for the PEG `a` - result.kind = pkCapturedSearch - result.sons = @[a] - -when false: - proc contains(a: TPeg, k: TPegKind): bool = - if a.kind == k: return true - case a.kind - of pkEmpty, pkAny, pkAnyRune, pkGreedyAny, pkNewLine, pkTerminal, - pkTerminalIgnoreCase, pkTerminalIgnoreStyle, pkChar, pkGreedyRepChar, - pkCharChoice, pkGreedyRepSet: discard - of pkNonTerminal: return true - else: - for i in 0..a.sons.len-1: - if contains(a.sons[i], k): return true - -proc `+`*(a: TPeg): TPeg {.rtl, extern: "npegsGreedyPosRep".} = - ## constructs a "greedy positive repetition" with the PEG `a` - return sequence(a, *a) - -proc `&`*(a: TPeg): TPeg {.rtl, extern: "npegsAndPredicate".} = - ## constructs an "and predicate" with the PEG `a` - result.kind = pkAndPredicate - result.sons = @[a] - -proc `!`*(a: TPeg): TPeg {.rtl, extern: "npegsNotPredicate".} = - ## constructs a "not predicate" with the PEG `a` - result.kind = pkNotPredicate - result.sons = @[a] - -proc any*: TPeg {.inline.} = - ## constructs the PEG `any character`:idx: (``.``) - result.kind = pkAny - -proc anyRune*: TPeg {.inline.} = - ## constructs the PEG `any rune`:idx: (``_``) - result.kind = pkAnyRune - -proc newLine*: TPeg {.inline.} = - ## constructs the PEG `newline`:idx: (``\n``) - result.kind = pkNewline - -proc UnicodeLetter*: TPeg {.inline.} = - ## constructs the PEG ``\letter`` which matches any Unicode letter. - result.kind = pkLetter - -proc UnicodeLower*: TPeg {.inline.} = - ## constructs the PEG ``\lower`` which matches any Unicode lowercase letter. - result.kind = pkLower - -proc UnicodeUpper*: TPeg {.inline.} = - ## constructs the PEG ``\upper`` which matches any Unicode lowercase letter. - result.kind = pkUpper - -proc UnicodeTitle*: TPeg {.inline.} = - ## constructs the PEG ``\title`` which matches any Unicode title letter. - result.kind = pkTitle - -proc UnicodeWhitespace*: TPeg {.inline.} = - ## constructs the PEG ``\white`` which matches any Unicode - ## whitespace character. - result.kind = pkWhitespace - -proc startAnchor*: TPeg {.inline.} = - ## constructs the PEG ``^`` which matches the start of the input. - result.kind = pkStartAnchor - -proc endAnchor*: TPeg {.inline.} = - ## constructs the PEG ``$`` which matches the end of the input. - result = !any() - -proc capture*(a: TPeg): TPeg {.rtl, extern: "npegsCapture".} = - ## constructs a capture with the PEG `a` - result.kind = pkCapture - result.sons = @[a] - -proc backref*(index: range[1..MaxSubPatterns]): TPeg {. - rtl, extern: "npegs$1".} = - ## constructs a back reference of the given `index`. `index` starts counting - ## from 1. - result.kind = pkBackRef - result.index = index-1 - -proc backrefIgnoreCase*(index: range[1..MaxSubPatterns]): TPeg {. - rtl, extern: "npegs$1".} = - ## constructs a back reference of the given `index`. `index` starts counting - ## from 1. Ignores case for matching. - result.kind = pkBackRefIgnoreCase - result.index = index-1 - -proc backrefIgnoreStyle*(index: range[1..MaxSubPatterns]): TPeg {. - rtl, extern: "npegs$1".}= - ## constructs a back reference of the given `index`. `index` starts counting - ## from 1. Ignores style for matching. - result.kind = pkBackRefIgnoreStyle - result.index = index-1 - -proc spaceCost(n: TPeg): int = - case n.kind - of pkEmpty: discard - of pkTerminal, pkTerminalIgnoreCase, pkTerminalIgnoreStyle, pkChar, - pkGreedyRepChar, pkCharChoice, pkGreedyRepSet, - pkAny..pkWhitespace, pkGreedyAny: - result = 1 - of pkNonTerminal: - # we cannot inline a rule with a non-terminal - result = InlineThreshold+1 - else: - for i in 0..n.len-1: - inc(result, spaceCost(n.sons[i])) - if result >= InlineThreshold: break - -proc nonterminal*(n: PNonTerminal): TPeg {. - rtl, extern: "npegs$1".} = - ## constructs a PEG that consists of the nonterminal symbol - assert n != nil - if ntDeclared in n.flags and spaceCost(n.rule) < InlineThreshold: - when false: echo "inlining symbol: ", n.name - result = n.rule # inlining of rule enables better optimizations - else: - result.kind = pkNonTerminal - result.nt = n - -proc newNonTerminal*(name: string, line, column: int): PNonTerminal {. - rtl, extern: "npegs$1".} = - ## constructs a nonterminal symbol - new(result) - result.name = name - result.line = line - result.col = column - -template letters*: TPeg = - ## expands to ``charset({'A'..'Z', 'a'..'z'})`` - charset({'A'..'Z', 'a'..'z'}) - -template digits*: TPeg = - ## expands to ``charset({'0'..'9'})`` - charset({'0'..'9'}) - -template whitespace*: TPeg = - ## expands to ``charset({' ', '\9'..'\13'})`` - charset({' ', '\9'..'\13'}) - -template identChars*: TPeg = - ## expands to ``charset({'a'..'z', 'A'..'Z', '0'..'9', '_'})`` - charset({'a'..'z', 'A'..'Z', '0'..'9', '_'}) - -template identStartChars*: TPeg = - ## expands to ``charset({'A'..'Z', 'a'..'z', '_'})`` - charset({'a'..'z', 'A'..'Z', '_'}) - -template ident*: TPeg = - ## same as ``[a-zA-Z_][a-zA-z_0-9]*``; standard identifier - sequence(charset({'a'..'z', 'A'..'Z', '_'}), - *charset({'a'..'z', 'A'..'Z', '0'..'9', '_'})) - -template natural*: TPeg = - ## same as ``\d+`` - +digits - -# ------------------------- debugging ----------------------------------------- - -proc esc(c: char, reserved = {'\0'..'\255'}): string = - case c - of '\b': result = "\\b" - of '\t': result = "\\t" - of '\c': result = "\\c" - of '\L': result = "\\l" - of '\v': result = "\\v" - of '\f': result = "\\f" - of '\e': result = "\\e" - of '\a': result = "\\a" - of '\\': result = "\\\\" - of 'a'..'z', 'A'..'Z', '0'..'9', '_': result = $c - elif c < ' ' or c >= '\128': result = '\\' & $ord(c) - elif c in reserved: result = '\\' & c - else: result = $c - -proc singleQuoteEsc(c: char): string = return "'" & esc(c, {'\''}) & "'" - -proc singleQuoteEsc(str: string): string = - result = "'" - for c in items(str): add result, esc(c, {'\''}) - add result, '\'' - -proc charSetEscAux(cc: set[char]): string = - const reserved = {'^', '-', ']'} - result = "" - var c1 = 0 - while c1 <= 0xff: - if chr(c1) in cc: - var c2 = c1 - while c2 < 0xff and chr(succ(c2)) in cc: inc(c2) - if c1 == c2: - add result, esc(chr(c1), reserved) - elif c2 == succ(c1): - add result, esc(chr(c1), reserved) & esc(chr(c2), reserved) - else: - add result, esc(chr(c1), reserved) & '-' & esc(chr(c2), reserved) - c1 = c2 - inc(c1) - -proc charSetEsc(cc: set[char]): string = - if card(cc) >= 128+64: - result = "[^" & charSetEscAux({'\1'..'\xFF'} - cc) & ']' - else: - result = '[' & charSetEscAux(cc) & ']' - -proc toStrAux(r: TPeg, res: var string) = - case r.kind - of pkEmpty: add(res, "()") - of pkAny: add(res, '.') - of pkAnyRune: add(res, '_') - of pkLetter: add(res, "\\letter") - of pkLower: add(res, "\\lower") - of pkUpper: add(res, "\\upper") - of pkTitle: add(res, "\\title") - of pkWhitespace: add(res, "\\white") - - of pkNewline: add(res, "\\n") - of pkTerminal: add(res, singleQuoteEsc(r.term)) - of pkTerminalIgnoreCase: - add(res, 'i') - add(res, singleQuoteEsc(r.term)) - of pkTerminalIgnoreStyle: - add(res, 'y') - add(res, singleQuoteEsc(r.term)) - of pkChar: add(res, singleQuoteEsc(r.ch)) - of pkCharChoice: add(res, charSetEsc(r.charChoice[])) - of pkNonTerminal: add(res, r.nt.name) - of pkSequence: - add(res, '(') - toStrAux(r.sons[0], res) - for i in 1 .. high(r.sons): - add(res, ' ') - toStrAux(r.sons[i], res) - add(res, ')') - of pkOrderedChoice: - add(res, '(') - toStrAux(r.sons[0], res) - for i in 1 .. high(r.sons): - add(res, " / ") - toStrAux(r.sons[i], res) - add(res, ')') - of pkGreedyRep: - toStrAux(r.sons[0], res) - add(res, '*') - of pkGreedyRepChar: - add(res, singleQuoteEsc(r.ch)) - add(res, '*') - of pkGreedyRepSet: - add(res, charSetEsc(r.charChoice[])) - add(res, '*') - of pkGreedyAny: - add(res, ".*") - of pkOption: - toStrAux(r.sons[0], res) - add(res, '?') - of pkAndPredicate: - add(res, '&') - toStrAux(r.sons[0], res) - of pkNotPredicate: - add(res, '!') - toStrAux(r.sons[0], res) - of pkSearch: - add(res, '@') - toStrAux(r.sons[0], res) - of pkCapturedSearch: - add(res, "{@}") - toStrAux(r.sons[0], res) - of pkCapture: - add(res, '{') - toStrAux(r.sons[0], res) - add(res, '}') - of pkBackRef: - add(res, '$') - add(res, $r.index) - of pkBackRefIgnoreCase: - add(res, "i$") - add(res, $r.index) - of pkBackRefIgnoreStyle: - add(res, "y$") - add(res, $r.index) - of pkRule: - toStrAux(r.sons[0], res) - add(res, " <- ") - toStrAux(r.sons[1], res) - of pkList: - for i in 0 .. high(r.sons): - toStrAux(r.sons[i], res) - add(res, "\n") - of pkStartAnchor: - add(res, '^') - -proc `$` *(r: TPeg): string {.rtl, extern: "npegsToString".} = - ## converts a PEG to its string representation - result = "" - toStrAux(r, result) - -# --------------------- core engine ------------------------------------------- - -type - TCaptures* {.final.} = object ## contains the captured substrings. - matches: array[0..MaxSubpatterns-1, tuple[first, last: int]] - ml: int - origStart: int - -proc bounds*(c: TCaptures, - i: range[0..MaxSubpatterns-1]): tuple[first, last: int] = - ## returns the bounds ``[first..last]`` of the `i`'th capture. - result = c.matches[i] - -when not useUnicode: - type - Rune = char - template fastRuneAt(s, i, ch) = - ch = s[i] - inc(i) - template runeLenAt(s, i): untyped = 1 - - proc isAlpha(a: char): bool {.inline.} = return a in {'a'..'z','A'..'Z'} - proc isUpper(a: char): bool {.inline.} = return a in {'A'..'Z'} - proc isLower(a: char): bool {.inline.} = return a in {'a'..'z'} - proc isTitle(a: char): bool {.inline.} = return false - proc isWhiteSpace(a: char): bool {.inline.} = return a in {' ', '\9'..'\13'} - -proc rawMatch*(s: string, p: TPeg, start: int, c: var TCaptures): int {. - rtl, extern: "npegs$1".} = - ## low-level matching proc that implements the PEG interpreter. Use this - ## for maximum efficiency (every other PEG operation ends up calling this - ## proc). - ## Returns -1 if it does not match, else the length of the match - case p.kind - of pkEmpty: result = 0 # match of length 0 - of pkAny: - if s[start] != '\0': result = 1 - else: result = -1 - of pkAnyRune: - if s[start] != '\0': - result = runeLenAt(s, start) - else: - result = -1 - of pkLetter: - if s[start] != '\0': - var a: Rune - result = start - fastRuneAt(s, result, a) - if isAlpha(a): dec(result, start) - else: result = -1 - else: - result = -1 - of pkLower: - if s[start] != '\0': - var a: Rune - result = start - fastRuneAt(s, result, a) - if isLower(a): dec(result, start) - else: result = -1 - else: - result = -1 - of pkUpper: - if s[start] != '\0': - var a: Rune - result = start - fastRuneAt(s, result, a) - if isUpper(a): dec(result, start) - else: result = -1 - else: - result = -1 - of pkTitle: - if s[start] != '\0': - var a: Rune - result = start - fastRuneAt(s, result, a) - if isTitle(a): dec(result, start) - else: result = -1 - else: - result = -1 - of pkWhitespace: - if s[start] != '\0': - var a: Rune - result = start - fastRuneAt(s, result, a) - if isWhitespace(a): dec(result, start) - else: result = -1 - else: - result = -1 - of pkGreedyAny: - result = len(s) - start - of pkNewLine: - if s[start] == '\L': result = 1 - elif s[start] == '\C': - if s[start+1] == '\L': result = 2 - else: result = 1 - else: result = -1 - of pkTerminal: - result = len(p.term) - for i in 0..result-1: - if p.term[i] != s[start+i]: - result = -1 - break - of pkTerminalIgnoreCase: - var - i = 0 - a, b: Rune - result = start - while i < len(p.term): - fastRuneAt(p.term, i, a) - fastRuneAt(s, result, b) - if toLower(a) != toLower(b): - result = -1 - break - dec(result, start) - of pkTerminalIgnoreStyle: - var - i = 0 - a, b: Rune - result = start - while i < len(p.term): - while true: - fastRuneAt(p.term, i, a) - if a != Rune('_'): break - while true: - fastRuneAt(s, result, b) - if b != Rune('_'): break - if toLower(a) != toLower(b): - result = -1 - break - dec(result, start) - of pkChar: - if p.ch == s[start]: result = 1 - else: result = -1 - of pkCharChoice: - if contains(p.charChoice[], s[start]): result = 1 - else: result = -1 - of pkNonTerminal: - var oldMl = c.ml - when false: echo "enter: ", p.nt.name - result = rawMatch(s, p.nt.rule, start, c) - when false: echo "leave: ", p.nt.name - if result < 0: c.ml = oldMl - of pkSequence: - var oldMl = c.ml - result = 0 - assert(not isNil(p.sons)) - for i in 0..high(p.sons): - var x = rawMatch(s, p.sons[i], start+result, c) - if x < 0: - c.ml = oldMl - result = -1 - break - else: inc(result, x) - of pkOrderedChoice: - var oldMl = c.ml - for i in 0..high(p.sons): - result = rawMatch(s, p.sons[i], start, c) - if result >= 0: break - c.ml = oldMl - of pkSearch: - var oldMl = c.ml - result = 0 - while start+result < s.len: - var x = rawMatch(s, p.sons[0], start+result, c) - if x >= 0: - inc(result, x) - return - inc(result) - result = -1 - c.ml = oldMl - of pkCapturedSearch: - var idx = c.ml # reserve a slot for the subpattern - inc(c.ml) - result = 0 - while start+result < s.len: - var x = rawMatch(s, p.sons[0], start+result, c) - if x >= 0: - if idx < MaxSubpatterns: - c.matches[idx] = (start, start+result-1) - #else: silently ignore the capture - inc(result, x) - return - inc(result) - result = -1 - c.ml = idx - of pkGreedyRep: - result = 0 - while true: - var x = rawMatch(s, p.sons[0], start+result, c) - # if x == 0, we have an endless loop; so the correct behaviour would be - # not to break. But endless loops can be easily introduced: - # ``(comment / \w*)*`` is such an example. Breaking for x == 0 does the - # expected thing in this case. - if x <= 0: break - inc(result, x) - of pkGreedyRepChar: - result = 0 - var ch = p.ch - while ch == s[start+result]: inc(result) - of pkGreedyRepSet: - result = 0 - while contains(p.charChoice[], s[start+result]): inc(result) - of pkOption: - result = max(0, rawMatch(s, p.sons[0], start, c)) - of pkAndPredicate: - var oldMl = c.ml - result = rawMatch(s, p.sons[0], start, c) - if result >= 0: result = 0 # do not consume anything - else: c.ml = oldMl - of pkNotPredicate: - var oldMl = c.ml - result = rawMatch(s, p.sons[0], start, c) - if result < 0: result = 0 - else: - c.ml = oldMl - result = -1 - of pkCapture: - var idx = c.ml # reserve a slot for the subpattern - inc(c.ml) - result = rawMatch(s, p.sons[0], start, c) - if result >= 0: - if idx < MaxSubpatterns: - c.matches[idx] = (start, start+result-1) - #else: silently ignore the capture - else: - c.ml = idx - of pkBackRef..pkBackRefIgnoreStyle: - if p.index >= c.ml: return -1 - var (a, b) = c.matches[p.index] - var n: TPeg - n.kind = succ(pkTerminal, ord(p.kind)-ord(pkBackRef)) - n.term = s.substr(a, b) - result = rawMatch(s, n, start, c) - of pkStartAnchor: - if c.origStart == start: result = 0 - else: result = -1 - of pkRule, pkList: assert false - -proc match*(s: string, pattern: TPeg, matches: var openarray[string], - start = 0): bool {.rtl, extern: "npegs$1Capture".} = - ## returns ``true`` if ``s[start..]`` matches the ``pattern`` and - ## the captured substrings in the array ``matches``. If it does not - ## match, nothing is written into ``matches`` and ``false`` is - ## returned. - var c: TCaptures - c.origStart = start - result = rawMatch(s, pattern, start, c) == len(s)-start - if result: - for i in 0..c.ml-1: - matches[i] = substr(s, c.matches[i][0], c.matches[i][1]) - -proc match*(s: string, pattern: TPeg, - start = 0): bool {.rtl, extern: "npegs$1".} = - ## returns ``true`` if ``s`` matches the ``pattern`` beginning from ``start``. - var c: TCaptures - c.origStart = start - result = rawMatch(s, pattern, start, c) == len(s)-start - -proc matchLen*(s: string, pattern: TPeg, matches: var openarray[string], - start = 0): int {.rtl, extern: "npegs$1Capture".} = - ## the same as ``match``, but it returns the length of the match, - ## if there is no match, -1 is returned. Note that a match length - ## of zero can happen. It's possible that a suffix of `s` remains - ## that does not belong to the match. - var c: TCaptures - c.origStart = start - result = rawMatch(s, pattern, start, c) - if result >= 0: - for i in 0..c.ml-1: - matches[i] = substr(s, c.matches[i][0], c.matches[i][1]) - -proc matchLen*(s: string, pattern: TPeg, - start = 0): int {.rtl, extern: "npegs$1".} = - ## the same as ``match``, but it returns the length of the match, - ## if there is no match, -1 is returned. Note that a match length - ## of zero can happen. It's possible that a suffix of `s` remains - ## that does not belong to the match. - var c: TCaptures - c.origStart = start - result = rawMatch(s, pattern, start, c) - -proc find*(s: string, pattern: TPeg, matches: var openarray[string], - start = 0): int {.rtl, extern: "npegs$1Capture".} = - ## returns the starting position of ``pattern`` in ``s`` and the captured - ## substrings in the array ``matches``. If it does not match, nothing - ## is written into ``matches`` and -1 is returned. - for i in start .. s.len-1: - if matchLen(s, pattern, matches, i) >= 0: return i - return -1 - # could also use the pattern here: (!P .)* P - -proc findBounds*(s: string, pattern: TPeg, matches: var openarray[string], - start = 0): tuple[first, last: int] {. - rtl, extern: "npegs$1Capture".} = - ## returns the starting position and end position of ``pattern`` in ``s`` - ## and the captured - ## substrings in the array ``matches``. If it does not match, nothing - ## is written into ``matches`` and (-1,0) is returned. - for i in start .. s.len-1: - var L = matchLen(s, pattern, matches, i) - if L >= 0: return (i, i+L-1) - return (-1, 0) - -proc find*(s: string, pattern: TPeg, - start = 0): int {.rtl, extern: "npegs$1".} = - ## returns the starting position of ``pattern`` in ``s``. If it does not - ## match, -1 is returned. - for i in start .. s.len-1: - if matchLen(s, pattern, i) >= 0: return i - return -1 - -iterator findAll*(s: string, pattern: TPeg, start = 0): string = - ## yields all matching captures of pattern in `s`. - var matches: array[0..MaxSubpatterns-1, string] - var i = start - while i < s.len: - var L = matchLen(s, pattern, matches, i) - if L < 0: break - for k in 0..MaxSubPatterns-1: - if isNil(matches[k]): break - yield matches[k] - inc(i, L) - -proc findAll*(s: string, pattern: TPeg, start = 0): seq[string] {. - rtl, extern: "npegs$1".} = - ## returns all matching captures of pattern in `s`. - ## If it does not match, @[] is returned. - accumulateResult(findAll(s, pattern, start)) - -template `=~`*(s: string, pattern: TPeg): untyped = - ## This calls ``match`` with an implicit declared ``matches`` array that - ## can be used in the scope of the ``=~`` call: - ## - ## .. code-block:: nim - ## - ## if line =~ peg"\s* {\w+} \s* '=' \s* {\w+}": - ## # matches a key=value pair: - ## echo("Key: ", matches[0]) - ## echo("Value: ", matches[1]) - ## elif line =~ peg"\s*{'#'.*}": - ## # matches a comment - ## # note that the implicit ``matches`` array is different from the - ## # ``matches`` array of the first branch - ## echo("comment: ", matches[0]) - ## else: - ## echo("syntax error") - ## - when not declaredInScope(matches): - var matches {.inject.}: array[0..MaxSubpatterns-1, string] - match(s, pattern, matches) - -# ------------------------- more string handling ------------------------------ - -proc contains*(s: string, pattern: TPeg, start = 0): bool {. - rtl, extern: "npegs$1".} = - ## same as ``find(s, pattern, start) >= 0`` - return find(s, pattern, start) >= 0 - -proc contains*(s: string, pattern: TPeg, matches: var openArray[string], - start = 0): bool {.rtl, extern: "npegs$1Capture".} = - ## same as ``find(s, pattern, matches, start) >= 0`` - return find(s, pattern, matches, start) >= 0 - -proc startsWith*(s: string, prefix: TPeg, start = 0): bool {. - rtl, extern: "npegs$1".} = - ## returns true if `s` starts with the pattern `prefix` - result = matchLen(s, prefix, start) >= 0 - -proc endsWith*(s: string, suffix: TPeg, start = 0): bool {. - rtl, extern: "npegs$1".} = - ## returns true if `s` ends with the pattern `prefix` - for i in start .. s.len-1: - if matchLen(s, suffix, i) == s.len - i: return true - -proc replacef*(s: string, sub: TPeg, by: string): string {. - rtl, extern: "npegs$1".} = - ## Replaces `sub` in `s` by the string `by`. Captures can be accessed in `by` - ## with the notation ``$i`` and ``$#`` (see strutils.`%`). Examples: - ## - ## .. code-block:: nim - ## "var1=key; var2=key2".replace(peg"{\ident}'='{\ident}", "$1<-$2$2") - ## - ## Results in: - ## - ## .. code-block:: nim - ## - ## "var1<-keykey; val2<-key2key2" - result = "" - var i = 0 - var caps: array[0..MaxSubpatterns-1, string] - while i < s.len: - var x = matchLen(s, sub, caps, i) - if x <= 0: - add(result, s[i]) - inc(i) - else: - addf(result, by, caps) - inc(i, x) - add(result, substr(s, i)) - -proc replace*(s: string, sub: TPeg, by = ""): string {. - rtl, extern: "npegs$1".} = - ## Replaces `sub` in `s` by the string `by`. Captures cannot be accessed - ## in `by`. - result = "" - var i = 0 - var caps: array[0..MaxSubpatterns-1, string] - while i < s.len: - var x = matchLen(s, sub, caps, i) - if x <= 0: - add(result, s[i]) - inc(i) - else: - addf(result, by, caps) - inc(i, x) - add(result, substr(s, i)) - -proc parallelReplace*(s: string, subs: varargs[ - tuple[pattern: TPeg, repl: string]]): string {. - rtl, extern: "npegs$1".} = - ## Returns a modified copy of `s` with the substitutions in `subs` - ## applied in parallel. - result = "" - var i = 0 - var caps: array[0..MaxSubpatterns-1, string] - while i < s.len: - block searchSubs: - for j in 0..high(subs): - var x = matchLen(s, subs[j][0], caps, i) - if x > 0: - addf(result, subs[j][1], caps) - inc(i, x) - break searchSubs - add(result, s[i]) - inc(i) - # copy the rest: - add(result, substr(s, i)) - -proc transformFile*(infile, outfile: string, - subs: varargs[tuple[pattern: TPeg, repl: string]]) {. - rtl, extern: "npegs$1".} = - ## reads in the file `infile`, performs a parallel replacement (calls - ## `parallelReplace`) and writes back to `outfile`. Calls ``quit`` if an - ## error occurs. This is supposed to be used for quick scripting. - var x = readFile(infile) - if not isNil(x): - var f: File - if open(f, outfile, fmWrite): - write(f, x.parallelReplace(subs)) - close(f) - else: - quit("cannot open for writing: " & outfile) - else: - quit("cannot open for reading: " & infile) - -iterator split*(s: string, sep: TPeg): string = - ## Splits the string `s` into substrings. - ## - ## Substrings are separated by the PEG `sep`. - ## Examples: - ## - ## .. code-block:: nim - ## for word in split("00232this02939is39an22example111", peg"\d+"): - ## writeLine(stdout, word) - ## - ## Results in: - ## - ## .. code-block:: nim - ## "this" - ## "is" - ## "an" - ## "example" - ## - var - first = 0 - last = 0 - while last < len(s): - var x = matchLen(s, sep, last) - if x > 0: inc(last, x) - first = last - while last < len(s): - inc(last) - x = matchLen(s, sep, last) - if x > 0: break - if first < last: - yield substr(s, first, last-1) - -proc split*(s: string, sep: TPeg): seq[string] {. - rtl, extern: "npegs$1".} = - ## Splits the string `s` into substrings. - accumulateResult(split(s, sep)) - -# ------------------- scanner ------------------------------------------------- - -type - TModifier = enum - modNone, - modVerbatim, - modIgnoreCase, - modIgnoreStyle - TTokKind = enum ## enumeration of all tokens - tkInvalid, ## invalid token - tkEof, ## end of file reached - tkAny, ## . - tkAnyRune, ## _ - tkIdentifier, ## abc - tkStringLit, ## "abc" or 'abc' - tkCharSet, ## [^A-Z] - tkParLe, ## '(' - tkParRi, ## ')' - tkCurlyLe, ## '{' - tkCurlyRi, ## '}' - tkCurlyAt, ## '{@}' - tkArrow, ## '<-' - tkBar, ## '/' - tkStar, ## '*' - tkPlus, ## '+' - tkAmp, ## '&' - tkNot, ## '!' - tkOption, ## '?' - tkAt, ## '@' - tkBuiltin, ## \identifier - tkEscaped, ## \\ - tkBackref, ## '$' - tkDollar, ## '$' - tkHat ## '^' - - TToken {.final.} = object ## a token - kind: TTokKind ## the type of the token - modifier: TModifier - literal: string ## the parsed (string) literal - charset: set[char] ## if kind == tkCharSet - index: int ## if kind == tkBackref - - TPegLexer {.inheritable.} = object ## the lexer object. - bufpos: int ## the current position within the buffer - buf: cstring ## the buffer itself - lineNumber: int ## the current line number - lineStart: int ## index of last line start in buffer - colOffset: int ## column to add - filename: string - -const - tokKindToStr: array[TTokKind, string] = [ - "invalid", "[EOF]", ".", "_", "identifier", "string literal", - "character set", "(", ")", "{", "}", "{@}", - "<-", "/", "*", "+", "&", "!", "?", - "@", "built-in", "escaped", "$", "$", "^" - ] - -proc HandleCR(L: var TPegLexer, pos: int): int = - assert(L.buf[pos] == '\c') - inc(L.linenumber) - result = pos+1 - if L.buf[result] == '\L': inc(result) - L.lineStart = result - -proc HandleLF(L: var TPegLexer, pos: int): int = - assert(L.buf[pos] == '\L') - inc(L.linenumber) - result = pos+1 - L.lineStart = result - -proc init(L: var TPegLexer, input, filename: string, line = 1, col = 0) = - L.buf = input - L.bufpos = 0 - L.lineNumber = line - L.colOffset = col - L.lineStart = 0 - L.filename = filename - -proc getColumn(L: TPegLexer): int {.inline.} = - result = abs(L.bufpos - L.lineStart) + L.colOffset - -proc getLine(L: TPegLexer): int {.inline.} = - result = L.linenumber - -proc errorStr(L: TPegLexer, msg: string, line = -1, col = -1): string = - var line = if line < 0: getLine(L) else: line - var col = if col < 0: getColumn(L) else: col - result = "$1($2, $3) Error: $4" % [L.filename, $line, $col, msg] - -proc handleHexChar(c: var TPegLexer, xi: var int) = - case c.buf[c.bufpos] - of '0'..'9': - xi = (xi shl 4) or (ord(c.buf[c.bufpos]) - ord('0')) - inc(c.bufpos) - of 'a'..'f': - xi = (xi shl 4) or (ord(c.buf[c.bufpos]) - ord('a') + 10) - inc(c.bufpos) - of 'A'..'F': - xi = (xi shl 4) or (ord(c.buf[c.bufpos]) - ord('A') + 10) - inc(c.bufpos) - else: discard - -proc getEscapedChar(c: var TPegLexer, tok: var TToken) = - inc(c.bufpos) - case c.buf[c.bufpos] - of 'r', 'R', 'c', 'C': - add(tok.literal, '\c') - inc(c.bufpos) - of 'l', 'L': - add(tok.literal, '\L') - inc(c.bufpos) - of 'f', 'F': - add(tok.literal, '\f') - inc(c.bufpos) - of 'e', 'E': - add(tok.literal, '\e') - inc(c.bufpos) - of 'a', 'A': - add(tok.literal, '\a') - inc(c.bufpos) - of 'b', 'B': - add(tok.literal, '\b') - inc(c.bufpos) - of 'v', 'V': - add(tok.literal, '\v') - inc(c.bufpos) - of 't', 'T': - add(tok.literal, '\t') - inc(c.bufpos) - of 'x', 'X': - inc(c.bufpos) - var xi = 0 - handleHexChar(c, xi) - handleHexChar(c, xi) - if xi == 0: tok.kind = tkInvalid - else: add(tok.literal, chr(xi)) - of '0'..'9': - var val = ord(c.buf[c.bufpos]) - ord('0') - inc(c.bufpos) - var i = 1 - while (i <= 3) and (c.buf[c.bufpos] in {'0'..'9'}): - val = val * 10 + ord(c.buf[c.bufpos]) - ord('0') - inc(c.bufpos) - inc(i) - if val > 0 and val <= 255: add(tok.literal, chr(val)) - else: tok.kind = tkInvalid - of '\0'..'\31': - tok.kind = tkInvalid - elif c.buf[c.bufpos] in strutils.Letters: - tok.kind = tkInvalid - else: - add(tok.literal, c.buf[c.bufpos]) - inc(c.bufpos) - -proc skip(c: var TPegLexer) = - var pos = c.bufpos - var buf = c.buf - while true: - case buf[pos] - of ' ', '\t': - inc(pos) - of '#': - while not (buf[pos] in {'\c', '\L', '\0'}): inc(pos) - of '\c': - pos = HandleCR(c, pos) - buf = c.buf - of '\L': - pos = HandleLF(c, pos) - buf = c.buf - else: - break # EndOfFile also leaves the loop - c.bufpos = pos - -proc getString(c: var TPegLexer, tok: var TToken) = - tok.kind = tkStringLit - var pos = c.bufPos + 1 - var buf = c.buf - var quote = buf[pos-1] - while true: - case buf[pos] - of '\\': - c.bufpos = pos - getEscapedChar(c, tok) - pos = c.bufpos - of '\c', '\L', '\0': - tok.kind = tkInvalid - break - elif buf[pos] == quote: - inc(pos) - break - else: - add(tok.literal, buf[pos]) - inc(pos) - c.bufpos = pos - -proc getDollar(c: var TPegLexer, tok: var TToken) = - var pos = c.bufPos + 1 - var buf = c.buf - if buf[pos] in {'0'..'9'}: - tok.kind = tkBackref - tok.index = 0 - while buf[pos] in {'0'..'9'}: - tok.index = tok.index * 10 + ord(buf[pos]) - ord('0') - inc(pos) - else: - tok.kind = tkDollar - c.bufpos = pos - -proc getCharSet(c: var TPegLexer, tok: var TToken) = - tok.kind = tkCharSet - tok.charset = {} - var pos = c.bufPos + 1 - var buf = c.buf - var caret = false - if buf[pos] == '^': - inc(pos) - caret = true - while true: - var ch: char - case buf[pos] - of ']': - inc(pos) - break - of '\\': - c.bufpos = pos - getEscapedChar(c, tok) - pos = c.bufpos - ch = tok.literal[tok.literal.len-1] - of '\C', '\L', '\0': - tok.kind = tkInvalid - break - else: - ch = buf[pos] - inc(pos) - incl(tok.charset, ch) - if buf[pos] == '-': - if buf[pos+1] == ']': - incl(tok.charset, '-') - inc(pos) - else: - inc(pos) - var ch2: char - case buf[pos] - of '\\': - c.bufpos = pos - getEscapedChar(c, tok) - pos = c.bufpos - ch2 = tok.literal[tok.literal.len-1] - of '\C', '\L', '\0': - tok.kind = tkInvalid - break - else: - ch2 = buf[pos] - inc(pos) - for i in ord(ch)+1 .. ord(ch2): - incl(tok.charset, chr(i)) - c.bufpos = pos - if caret: tok.charset = {'\1'..'\xFF'} - tok.charset - -proc getSymbol(c: var TPegLexer, tok: var TToken) = - var pos = c.bufpos - var buf = c.buf - while true: - add(tok.literal, buf[pos]) - inc(pos) - if buf[pos] notin strutils.IdentChars: break - c.bufpos = pos - tok.kind = tkIdentifier - -proc getBuiltin(c: var TPegLexer, tok: var TToken) = - if c.buf[c.bufpos+1] in strutils.Letters: - inc(c.bufpos) - getSymbol(c, tok) - tok.kind = tkBuiltin - else: - tok.kind = tkEscaped - getEscapedChar(c, tok) # may set tok.kind to tkInvalid - -proc getTok(c: var TPegLexer, tok: var TToken) = - tok.kind = tkInvalid - tok.modifier = modNone - setlen(tok.literal, 0) - skip(c) - case c.buf[c.bufpos] - of '{': - inc(c.bufpos) - if c.buf[c.bufpos] == '@' and c.buf[c.bufpos+1] == '}': - tok.kind = tkCurlyAt - inc(c.bufpos, 2) - add(tok.literal, "{@}") - else: - tok.kind = tkCurlyLe - add(tok.literal, '{') - of '}': - tok.kind = tkCurlyRi - inc(c.bufpos) - add(tok.literal, '}') - of '[': - getCharset(c, tok) - of '(': - tok.kind = tkParLe - inc(c.bufpos) - add(tok.literal, '(') - of ')': - tok.kind = tkParRi - inc(c.bufpos) - add(tok.literal, ')') - of '.': - tok.kind = tkAny - inc(c.bufpos) - add(tok.literal, '.') - of '_': - tok.kind = tkAnyRune - inc(c.bufpos) - add(tok.literal, '_') - of '\\': - getBuiltin(c, tok) - of '\'', '"': getString(c, tok) - of '$': getDollar(c, tok) - of '\0': - tok.kind = tkEof - tok.literal = "[EOF]" - of 'a'..'z', 'A'..'Z', '\128'..'\255': - getSymbol(c, tok) - if c.buf[c.bufpos] in {'\'', '"'} or - c.buf[c.bufpos] == '$' and c.buf[c.bufpos+1] in {'0'..'9'}: - case tok.literal - of "i": tok.modifier = modIgnoreCase - of "y": tok.modifier = modIgnoreStyle - of "v": tok.modifier = modVerbatim - else: discard - setLen(tok.literal, 0) - if c.buf[c.bufpos] == '$': - getDollar(c, tok) - else: - getString(c, tok) - if tok.modifier == modNone: tok.kind = tkInvalid - of '+': - tok.kind = tkPlus - inc(c.bufpos) - add(tok.literal, '+') - of '*': - tok.kind = tkStar - inc(c.bufpos) - add(tok.literal, '+') - of '<': - if c.buf[c.bufpos+1] == '-': - inc(c.bufpos, 2) - tok.kind = tkArrow - add(tok.literal, "<-") - else: - add(tok.literal, '<') - of '/': - tok.kind = tkBar - inc(c.bufpos) - add(tok.literal, '/') - of '?': - tok.kind = tkOption - inc(c.bufpos) - add(tok.literal, '?') - of '!': - tok.kind = tkNot - inc(c.bufpos) - add(tok.literal, '!') - of '&': - tok.kind = tkAmp - inc(c.bufpos) - add(tok.literal, '!') - of '@': - tok.kind = tkAt - inc(c.bufpos) - add(tok.literal, '@') - if c.buf[c.bufpos] == '@': - tok.kind = tkCurlyAt - inc(c.bufpos) - add(tok.literal, '@') - of '^': - tok.kind = tkHat - inc(c.bufpos) - add(tok.literal, '^') - else: - add(tok.literal, c.buf[c.bufpos]) - inc(c.bufpos) - -proc arrowIsNextTok(c: TPegLexer): bool = - # the only look ahead we need - var pos = c.bufpos - while c.buf[pos] in {'\t', ' '}: inc(pos) - result = c.buf[pos] == '<' and c.buf[pos+1] == '-' - -# ----------------------------- parser ---------------------------------------- - -type - EInvalidPeg* = object of ValueError ## raised if an invalid - ## PEG has been detected - TPegParser = object of TPegLexer ## the PEG parser object - tok: TToken - nonterms: seq[PNonTerminal] - modifier: TModifier - captures: int - identIsVerbatim: bool - skip: TPeg - -proc pegError(p: TPegParser, msg: string, line = -1, col = -1) = - var e: ref EInvalidPeg - new(e) - e.msg = errorStr(p, msg, line, col) - raise e - -proc getTok(p: var TPegParser) = - getTok(p, p.tok) - if p.tok.kind == tkInvalid: pegError(p, "invalid token") - -proc eat(p: var TPegParser, kind: TTokKind) = - if p.tok.kind == kind: getTok(p) - else: pegError(p, tokKindToStr[kind] & " expected") - -proc parseExpr(p: var TPegParser): TPeg - -proc getNonTerminal(p: var TPegParser, name: string): PNonTerminal = - for i in 0..high(p.nonterms): - result = p.nonterms[i] - if cmpIgnoreStyle(result.name, name) == 0: return - # forward reference: - result = newNonTerminal(name, getLine(p), getColumn(p)) - add(p.nonterms, result) - -proc modifiedTerm(s: string, m: TModifier): TPeg = - case m - of modNone, modVerbatim: result = term(s) - of modIgnoreCase: result = termIgnoreCase(s) - of modIgnoreStyle: result = termIgnoreStyle(s) - -proc modifiedBackref(s: int, m: TModifier): TPeg = - case m - of modNone, modVerbatim: result = backRef(s) - of modIgnoreCase: result = backRefIgnoreCase(s) - of modIgnoreStyle: result = backRefIgnoreStyle(s) - -proc builtin(p: var TPegParser): TPeg = - # do not use "y", "skip" or "i" as these would be ambiguous - case p.tok.literal - of "n": result = newLine() - of "d": result = charset({'0'..'9'}) - of "D": result = charset({'\1'..'\xff'} - {'0'..'9'}) - of "s": result = charset({' ', '\9'..'\13'}) - of "S": result = charset({'\1'..'\xff'} - {' ', '\9'..'\13'}) - of "w": result = charset({'a'..'z', 'A'..'Z', '_', '0'..'9'}) - of "W": result = charset({'\1'..'\xff'} - {'a'..'z','A'..'Z','_','0'..'9'}) - of "a": result = charset({'a'..'z', 'A'..'Z'}) - of "A": result = charset({'\1'..'\xff'} - {'a'..'z', 'A'..'Z'}) - of "ident": result = tpegs.ident - of "letter": result = UnicodeLetter() - of "upper": result = UnicodeUpper() - of "lower": result = UnicodeLower() - of "title": result = UnicodeTitle() - of "white": result = UnicodeWhitespace() - else: pegError(p, "unknown built-in: " & p.tok.literal) - -proc token(terminal: TPeg, p: TPegParser): TPeg = - if p.skip.kind == pkEmpty: result = terminal - else: result = sequence(p.skip, terminal) - -proc primary(p: var TPegParser): TPeg = - case p.tok.kind - of tkAmp: - getTok(p) - return &primary(p) - of tkNot: - getTok(p) - return !primary(p) - of tkAt: - getTok(p) - return !*primary(p) - of tkCurlyAt: - getTok(p) - return !*\primary(p).token(p) - else: discard - case p.tok.kind - of tkIdentifier: - if p.identIsVerbatim: - var m = p.tok.modifier - if m == modNone: m = p.modifier - result = modifiedTerm(p.tok.literal, m).token(p) - getTok(p) - elif not arrowIsNextTok(p): - var nt = getNonTerminal(p, p.tok.literal) - incl(nt.flags, ntUsed) - result = nonTerminal(nt).token(p) - getTok(p) - else: - pegError(p, "expression expected, but found: " & p.tok.literal) - of tkStringLit: - var m = p.tok.modifier - if m == modNone: m = p.modifier - result = modifiedTerm(p.tok.literal, m).token(p) - getTok(p) - of tkCharSet: - if '\0' in p.tok.charset: - pegError(p, "binary zero ('\\0') not allowed in character class") - result = charset(p.tok.charset).token(p) - getTok(p) - of tkParLe: - getTok(p) - result = parseExpr(p) - eat(p, tkParRi) - of tkCurlyLe: - getTok(p) - result = capture(parseExpr(p)).token(p) - eat(p, tkCurlyRi) - inc(p.captures) - of tkAny: - result = any().token(p) - getTok(p) - of tkAnyRune: - result = anyRune().token(p) - getTok(p) - of tkBuiltin: - result = builtin(p).token(p) - getTok(p) - of tkEscaped: - result = term(p.tok.literal[0]).token(p) - getTok(p) - of tkDollar: - result = endAnchor() - getTok(p) - of tkHat: - result = startAnchor() - getTok(p) - of tkBackref: - var m = p.tok.modifier - if m == modNone: m = p.modifier - result = modifiedBackRef(p.tok.index, m).token(p) - if p.tok.index < 0 or p.tok.index > p.captures: - pegError(p, "invalid back reference index: " & $p.tok.index) - getTok(p) - else: - pegError(p, "expression expected, but found: " & p.tok.literal) - getTok(p) # we must consume a token here to prevent endless loops! - while true: - case p.tok.kind - of tkOption: - result = ?result - getTok(p) - of tkStar: - result = *result - getTok(p) - of tkPlus: - result = +result - getTok(p) - else: break - -proc seqExpr(p: var TPegParser): TPeg = - result = primary(p) - while true: - case p.tok.kind - of tkAmp, tkNot, tkAt, tkStringLit, tkCharset, tkParLe, tkCurlyLe, - tkAny, tkAnyRune, tkBuiltin, tkEscaped, tkDollar, tkBackref, - tkHat, tkCurlyAt: - result = sequence(result, primary(p)) - of tkIdentifier: - if not arrowIsNextTok(p): - result = sequence(result, primary(p)) - else: break - else: break - -proc parseExpr(p: var TPegParser): TPeg = - result = seqExpr(p) - while p.tok.kind == tkBar: - getTok(p) - result = result / seqExpr(p) - -proc parseRule(p: var TPegParser): PNonTerminal = - if p.tok.kind == tkIdentifier and arrowIsNextTok(p): - result = getNonTerminal(p, p.tok.literal) - if ntDeclared in result.flags: - pegError(p, "attempt to redefine: " & result.name) - result.line = getLine(p) - result.col = getColumn(p) - getTok(p) - eat(p, tkArrow) - result.rule = parseExpr(p) - incl(result.flags, ntDeclared) # NOW inlining may be attempted - else: - pegError(p, "rule expected, but found: " & p.tok.literal) - -proc rawParse(p: var TPegParser): TPeg = - ## parses a rule or a PEG expression - while p.tok.kind == tkBuiltin: - case p.tok.literal - of "i": - p.modifier = modIgnoreCase - getTok(p) - of "y": - p.modifier = modIgnoreStyle - getTok(p) - of "skip": - getTok(p) - p.skip = ?primary(p) - else: break - if p.tok.kind == tkIdentifier and arrowIsNextTok(p): - result = parseRule(p).rule - while p.tok.kind != tkEof: - discard parseRule(p) - else: - p.identIsVerbatim = true - result = parseExpr(p) - if p.tok.kind != tkEof: - pegError(p, "EOF expected, but found: " & p.tok.literal) - for i in 0..high(p.nonterms): - var nt = p.nonterms[i] - if ntDeclared notin nt.flags: - pegError(p, "undeclared identifier: " & nt.name, nt.line, nt.col) - elif ntUsed notin nt.flags and i > 0: - pegError(p, "unused rule: " & nt.name, nt.line, nt.col) - -proc parsePeg*(pattern: string, filename = "pattern", line = 1, col = 0): TPeg = - ## constructs a TPeg object from `pattern`. `filename`, `line`, `col` are - ## used for error messages, but they only provide start offsets. `parsePeg` - ## keeps track of line and column numbers within `pattern`. - var p: TPegParser - init(TPegLexer(p), pattern, filename, line, col) - p.tok.kind = tkInvalid - p.tok.modifier = modNone - p.tok.literal = "" - p.tok.charset = {} - p.nonterms = @[] - p.identIsVerbatim = false - getTok(p) - result = rawParse(p) - -proc peg*(pattern: string): TPeg = - ## constructs a TPeg object from the `pattern`. The short name has been - ## chosen to encourage its use as a raw string modifier:: - ## - ## peg"{\ident} \s* '=' \s* {.*}" - result = parsePeg(pattern, "pattern") - -proc escapePeg*(s: string): string = - ## escapes `s` so that it is matched verbatim when used as a peg. - result = "" - var inQuote = false - for c in items(s): - case c - of '\0'..'\31', '\'', '"', '\\': - if inQuote: - result.add('\'') - inQuote = false - result.add("\\x") - result.add(toHex(ord(c), 2)) - else: - if not inQuote: - result.add('\'') - inQuote = true - result.add(c) - if inQuote: result.add('\'') - -when isMainModule: - doAssert escapePeg("abc''def'") == r"'abc'\x27\x27'def'\x27" - #doAssert match("(a b c)", peg"'(' @ ')'") - doAssert match("W_HI_Le", peg"\y 'while'") - doAssert(not match("W_HI_L", peg"\y 'while'")) - doAssert(not match("W_HI_Le", peg"\y v'while'")) - doAssert match("W_HI_Le", peg"y'while'") - - doAssert($ +digits == $peg"\d+") - doAssert "0158787".match(peg"\d+") - doAssert "ABC 0232".match(peg"\w+\s+\d+") - doAssert "ABC".match(peg"\d+ / \w+") - - for word in split("00232this02939is39an22example111", peg"\d+"): - writeLine(stdout, word) - - doAssert matchLen("key", ident) == 3 - - var pattern = sequence(ident, *whitespace, term('='), *whitespace, ident) - doAssert matchLen("key1= cal9", pattern) == 11 - - var ws = newNonTerminal("ws", 1, 1) - ws.rule = *whitespace - - var expr = newNonTerminal("expr", 1, 1) - expr.rule = sequence(capture(ident), *sequence( - nonterminal(ws), term('+'), nonterminal(ws), nonterminal(expr))) - - var c: TCaptures - var s = "a+b + c +d+e+f" - doAssert rawMatch(s, expr.rule, 0, c) == len(s) - var a = "" - for i in 0..c.ml-1: - a.add(substr(s, c.matches[i][0], c.matches[i][1])) - doAssert a == "abcdef" - #echo expr.rule - - #const filename = "lib/devel/peg/grammar.txt" - #var grammar = parsePeg(newFileStream(filename, fmRead), filename) - #echo "a <- [abc]*?".match(grammar) - doAssert find("_____abc_______", term("abc"), 2) == 5 - doAssert match("_______ana", peg"A <- 'ana' / . A") - doAssert match("abcs%%%", peg"A <- ..A / .A / '%'") - - if "abc" =~ peg"{'a'}'bc' 'xyz' / {\ident}": - doAssert matches[0] == "abc" - else: - doAssert false - - var g2 = peg"""S <- A B / C D - A <- 'a'+ - B <- 'b'+ - C <- 'c'+ - D <- 'd'+ - """ - doAssert($g2 == "((A B) / (C D))") - doAssert match("cccccdddddd", g2) - doAssert("var1=key; var2=key2".replacef(peg"{\ident}'='{\ident}", "$1<-$2$2") == - "var1<-keykey; var2<-key2key2") - doAssert "var1=key; var2=key2".endsWith(peg"{\ident}'='{\ident}") - - if "aaaaaa" =~ peg"'aa' !. / ({'a'})+": - doAssert matches[0] == "a" - else: - doAssert false - - block: - var matches: array[0..2, string] - if match("abcdefg", peg"c {d} ef {g}", matches, 2): - doAssert matches[0] == "d" - doAssert matches[1] == "g" - else: - doAssert false - - for x in findAll("abcdef", peg"{.}", 3): - echo x - - if "f(a, b)" =~ peg"{[0-9]+} / ({\ident} '(' {@} ')')": - doAssert matches[0] == "f" - doAssert matches[1] == "a, b" - else: - doAssert false - - doAssert match("eine übersicht und außerdem", peg"(\letter \white*)+") - # ß is not a lower cased letter?! - doAssert match("eine übersicht und auerdem", peg"(\lower \white*)+") - doAssert match("EINE ÜBERSICHT UND AUSSERDEM", peg"(\upper \white*)+") - doAssert(not match("456678", peg"(\letter)+")) - - doAssert("var1 = key; var2 = key2".replacef( - peg"\skip(\s*) {\ident}'='{\ident}", "$1<-$2$2") == - "var1<-keykey;var2<-key2key2") - - doAssert match("prefix/start", peg"^start$", 7) - - # tricky test to check for false aliasing: - block: - var a = term"key" - echo($sequence(sequence(a, term"value"), *a)) - diff --git a/tests/stdlib/twchartoutf8.nim b/tests/stdlib/twchartoutf8.nim index b2f68ee32..a6602e3e3 100644 --- a/tests/stdlib/twchartoutf8.nim +++ b/tests/stdlib/twchartoutf8.nim @@ -30,7 +30,6 @@ else: result = newString(size) let res = WideCharToMultiByte(CP_UTF8, 0'i32, cast[LPWCSTR](addr(wc[0])), wclen, cstring(result), size, cstring(nil), LPBOOL(nil)) - result[size] = chr(0) doAssert size == res proc testCP(wc: WideCString, lo, hi: int) = diff --git a/tests/system/tnilconcats.nim b/tests/system/tnilconcats.nim new file mode 100644 index 000000000..ce059b7b0 --- /dev/null +++ b/tests/system/tnilconcats.nim @@ -0,0 +1,25 @@ +discard """ + output: '''@[nil, nil, nil, nil, nil, nil, nil, "meh"]''' + exitcode: "0" +""" + +when true: + var ab: string + ab &= "more" + + doAssert ab == "more" + + var x: seq[string] + + setLen(x, 7) + + x.add "meh" + + var s: string + var z = "abc" + var zz: string + s &= "foo" & z & zz + + doAssert s == "fooabc" + + echo x diff --git a/tests/system/toString.nim b/tests/system/toString.nim index ea9d6b05b..37c678a74 100644 --- a/tests/system/toString.nim +++ b/tests/system/toString.nim @@ -51,3 +51,59 @@ import strutils let arr = ['H','e','l','l','o',' ','W','o','r','l','d','!','\0'] doAssert $arr == "['H', 'e', 'l', 'l', 'o', ' ', 'W', 'o', 'r', 'l', 'd', '!', '\\x00']" doAssert $cstring(unsafeAddr arr) == "Hello World!" + +proc takes(c: cstring) = + doAssert c == "" + +proc testm() = + var x: string + # nil is mapped to "": + takes(x) + +testm() + +# nil tests +var xx: seq[string] +var yy: string +doAssert xx == @[] +doAssert yy == "" + +proc bar(arg: cstring): void = + doAssert arg[0] == '\0' + +proc baz(arg: openarray[char]): void = + doAssert arg.len == 0 + +proc stringCompare(): void = + var a,b,c,d,e,f,g: string + a.add 'a' + doAssert a == "a" + b.add "bee" + doAssert b == "bee" + b.add g + doAssert b == "bee" + c.add 123.456 + doAssert c == "123.456" + d.add 123456 + doAssert d == "123456" + + doAssert e == "" + doAssert "" == e + doAssert nil == e + doAssert e == nil + doAssert f == g + doAssert "" == "" + doAssert "" == nil + doAssert nil == "" + + g.setLen(10) + doAssert g == "\0\0\0\0\0\0\0\0\0\0" + doAssert "" != "\0\0\0\0\0\0\0\0\0\0" + + var nilstring: string + bar(nilstring) + baz(nilstring) + +stringCompare() +static: + stringCompare() \ No newline at end of file diff --git a/tests/typerel/t7600_1.nim b/tests/typerel/t7600_1.nim new file mode 100644 index 000000000..e3a5fefa2 --- /dev/null +++ b/tests/typerel/t7600_1.nim @@ -0,0 +1,18 @@ +discard """ +errormsg: "type mismatch: got <Thin[system.int]>" +nimout: '''t7600_1.nim(18, 6) Error: type mismatch: got <Thin[system.int]> +but expected one of: +proc test[T](x: Paper[T]) + +expression: test tn''' +""" + +type + Paper[T] = ref object of RootObj + thickness: T + Thin[T] = object of Paper[T] + +proc test[T](x: Paper[T]) = discard + +var tn = Thin[int]() +test tn diff --git a/tests/typerel/t7600_2.nim b/tests/typerel/t7600_2.nim new file mode 100644 index 000000000..7badb69cf --- /dev/null +++ b/tests/typerel/t7600_2.nim @@ -0,0 +1,17 @@ +discard """ +errormsg: "type mismatch: got <Thin>" +nimout: '''t7600_2.nim(17, 6) Error: type mismatch: got <Thin> +but expected one of: +proc test(x: Paper) + +expression: test tn''' +""" + +type + Paper = ref object of RootObj + Thin = object of Paper + +proc test(x: Paper) = discard + +var tn = Thin() +test tn diff --git a/tests/typerel/ttypelessemptyset.nim b/tests/typerel/ttypelessemptyset.nim index 3e171387b..5f49c33fd 100644 --- a/tests/typerel/ttypelessemptyset.nim +++ b/tests/typerel/ttypelessemptyset.nim @@ -1,5 +1,5 @@ discard """ - errormsg: "internal error: invalid kind for last(tyEmpty)" + errormsg: "internal error: invalid kind for lastOrd(tyEmpty)" """ var q = false discard (if q: {} else: {}) |