diff options
Diffstat (limited to 'lib/packages/docutils/rstast.nim')
| -rw-r--r-- | lib/packages/docutils/rstast.nim | 195 |
1 files changed, 163 insertions, 32 deletions
diff --git a/lib/packages/docutils/rstast.nim b/lib/packages/docutils/rstast.nim index f3596b571..2bbb0d0b8 100644 --- a/lib/packages/docutils/rstast.nim +++ b/lib/packages/docutils/rstast.nim @@ -9,20 +9,25 @@ ## This module implements an AST for the `reStructuredText`:idx: parser. -import strutils, json +import std/[strutils, json] + +when defined(nimPreviewSlimSystem): + import std/assertions + type RstNodeKind* = enum ## the possible node kinds of an PRstNode rnInner, # an inner node or a root rnHeadline, # a headline rnOverline, # an over- and underlined headline + rnMarkdownHeadline, # a Markdown headline rnTransition, # a transition (the ------------- <hr> thingie) rnParagraph, # a paragraph rnBulletList, # a bullet list rnBulletItem, # a bullet item rnEnumList, # an enumerated list rnEnumItem, # an enumerated item - rnDefList, # a definition list + rnDefList, rnMdDefList, # a definition list (RST/Markdown) rnDefItem, # an item of a definition list consisting of ... rnDefName, # ... a name part ... rnDefBody, # ... and a body part ... @@ -31,16 +36,28 @@ type rnFieldName, # consisting of a field name ... rnFieldBody, # ... and a field body rnOptionList, rnOptionListItem, rnOptionGroup, rnOption, rnOptionString, - rnOptionArgument, rnDescription, rnLiteralBlock, rnQuotedLiteralBlock, + rnOptionArgument, rnDescription, rnLiteralBlock, + rnMarkdownBlockQuote, # a quote starting from punctuation like >>> + rnMarkdownBlockQuoteItem, # a quotation block, quote lines starting with + # the same number of chars rnLineBlock, # the | thingie - rnLineBlockItem, # sons of the | thing + rnLineBlockItem, # a son of rnLineBlock - one line inside it. + # When `RstNode` lineIndent="\n" the line's empty rnBlockQuote, # text just indented - rnTable, rnGridTable, rnTableRow, rnTableHeaderCell, rnTableDataCell, - rnLabel, # used for footnotes and other things + rnTable, rnGridTable, rnMarkdownTable, rnTableRow, rnTableHeaderCell, rnTableDataCell, rnFootnote, # a footnote - rnCitation, # similar to footnote - rnStandaloneHyperlink, rnHyperlink, rnRef, rnDirective, # a directive - rnDirArg, rnRaw, rnTitle, rnContents, rnImage, rnFigure, rnCodeBlock, + rnCitation, # similar to footnote, so use rnFootnote instead + rnFootnoteGroup, # footnote group - exists for a purely stylistic + # reason: to display a few footnotes as 1 block + rnStandaloneHyperlink, rnHyperlink, + rnRstRef, # RST reference like `section name`_ + rnPandocRef, # Pandoc Markdown reference like [section name] + rnInternalRef, rnFootnoteRef, + rnNimdocRef, # reference to automatically generated Nim symbol + rnDirective, # a general directive + rnDirArg, # a directive argument (for some directives). + # here are directives that are not rnDirective: + rnRaw, rnTitle, rnContents, rnImage, rnFigure, rnCodeBlock, rnAdmonition, rnRawHtml, rnRawLatex, rnContainer, # ``container`` directive rnIndex, # index directve: @@ -49,46 +66,103 @@ type # * `file#id <file#id>`_ # * `file#id <file#id>'_ rnSubstitutionDef, # a definition of a substitution - rnGeneralRole, # Inline markup: + # Inline markup: + rnInlineCode, # interpreted text with code in a known language + rnCodeFragment, # inline code for highlighting with the specified + # class (which cannot be inferred from context) + rnUnknownRole, # interpreted text with an unknown role rnSub, rnSup, rnIdx, rnEmphasis, # "*" rnStrongEmphasis, # "**" rnTripleEmphasis, # "***" - rnInterpretedText, # "`" + rnInterpretedText, # "`" an auxiliary role for parsing that will + # be converted into other kinds like rnInlineCode rnInlineLiteral, # "``" + rnInlineTarget, # "_`target`" rnSubstitutionReferences, # "|" rnSmiley, # some smiley + rnDefaultRole, # .. default-role:: code rnLeaf # a leaf; the node's text field contains the # leaf val + FileIndex* = distinct int32 + TLineInfo* = object + line*: uint16 + col*: int16 + fileIndex*: FileIndex PRstNode* = ref RstNode ## an RST node RstNodeSeq* = seq[PRstNode] - RstNode* {.acyclic, final.} = object ## an RST node's description - kind*: RstNodeKind ## the node's kind - text*: string ## valid for leafs in the AST; and the title of - ## the document or the section - level*: int ## valid for some node kinds + RstNode* {.acyclic, final.} = object ## AST node (result of RST parsing) + case kind*: RstNodeKind ## the node's kind + of rnLeaf, rnSmiley: + text*: string ## string that is expected to be displayed + of rnEnumList: + labelFmt*: string ## label format like "(1)" + of rnLineBlockItem: + lineIndent*: string ## a few spaces or newline at the line beginning + of rnAdmonition: + adType*: string ## admonition type: "note", "caution", etc. This + ## text will set the style and also be displayed + of rnOverline, rnHeadline, rnMarkdownHeadline: + level*: int ## level of headings starting from 1 (main + ## chapter) to larger ones (minor sub-sections) + ## level=0 means it's document title or subtitle + of rnFootnote, rnCitation, rnOptionListItem: + order*: int ## footnote order (for auto-symbol footnotes and + ## auto-numbered ones without a label) + of rnMarkdownBlockQuoteItem: + quotationDepth*: int ## number of characters in line prefix + of rnRstRef, rnPandocRef, rnSubstitutionReferences, + rnInterpretedText, rnField, rnInlineCode, rnCodeBlock, rnFootnoteRef: + info*: TLineInfo ## To have line/column info for warnings at + ## nodes that are post-processed after parsing + of rnNimdocRef: + tooltip*: string + of rnTable, rnGridTable, rnMarkdownTable: + colCount*: int ## Number of (not-united) cells in the table + of rnTableRow: + endsHeader*: bool ## Is last row in the header of table? + of rnTableHeaderCell, rnTableDataCell: + span*: int ## Number of table columns that the cell occupies + else: + discard + anchor*: string ## anchor, internal link target + ## (aka HTML id tag, aka Latex label/hypertarget) sons*: RstNodeSeq ## the node's sons +proc `==`*(a, b: FileIndex): bool {.borrow.} + proc len*(n: PRstNode): int = result = len(n.sons) -proc newRstNode*(kind: RstNodeKind): PRstNode = - new(result) - result.sons = @[] - result.kind = kind +proc newRstNode*(kind: RstNodeKind, sons: seq[PRstNode] = @[], + anchor = ""): PRstNode = + result = PRstNode(kind: kind, sons: sons, anchor: anchor) -proc newRstNode*(kind: RstNodeKind, s: string): PRstNode = +proc newRstNode*(kind: RstNodeKind, info: TLineInfo, + sons: seq[PRstNode] = @[]): PRstNode = + result = PRstNode(kind: kind, sons: sons) + result.info = info + +proc newRstNode*(kind: RstNodeKind, s: string): PRstNode {.deprecated.} = + assert kind in {rnLeaf, rnSmiley} result = newRstNode(kind) result.text = s +proc newRstLeaf*(s: string): PRstNode = + result = newRstNode(rnLeaf) + result.text = s + proc lastSon*(n: PRstNode): PRstNode = result = n.sons[len(n.sons)-1] proc add*(father, son: PRstNode) = add(father.sons, son) +proc add*(father: PRstNode; s: string) = + add(father.sons, newRstLeaf(s)) + proc addIfNotNil*(father, son: PRstNode) = if son != nil: add(father, son) @@ -210,7 +284,7 @@ proc renderRstToRst(d: var RenderContext, n: PRstNode, result: var string) = inc(d.indent, 2) renderRstSons(d, n, result) dec(d.indent, 2) - of rnRef: + of rnRstRef: result.add("`") renderRstSons(d, n, result) result.add("`_") @@ -220,7 +294,7 @@ proc renderRstToRst(d: var RenderContext, n: PRstNode, result: var string) = result.add(" <") renderRstToRst(d, n.sons[1], result) result.add(">`_") - of rnGeneralRole: + of rnUnknownRole: result.add('`') renderRstToRst(d, n.sons[0],result) result.add("`:") @@ -293,9 +367,9 @@ proc renderRstToJsonNode(node: PRstNode): JsonNode = (key: "kind", val: %($node.kind)), (key: "level", val: %BiggestInt(node.level)) ] - if node.text != nil: + if node.kind in {rnLeaf, rnSmiley} and node.text.len > 0: result.add("text", %node.text) - if node.sons != nil and len(node.sons) > 0: + if len(node.sons) > 0: var accm = newSeq[JsonNode](len(node.sons)) for i, son in node.sons: accm[i] = renderRstToJsonNode(son) @@ -303,11 +377,68 @@ proc renderRstToJsonNode(node: PRstNode): JsonNode = proc renderRstToJson*(node: PRstNode): string = ## Writes the given RST node as JSON that is in the form - ## :: - ## { - ## "kind":string node.kind, - ## "text":optional string node.text, - ## "level":optional int node.level, - ## "sons":optional node array - ## } + ## + ## { + ## "kind":string node.kind, + ## "text":optional string node.text, + ## "level":optional int node.level, + ## "sons":optional node array + ## } renderRstToJsonNode(node).pretty + +proc renderRstToText*(node: PRstNode): string = + ## minimal text representation of markup node + const code = {rnCodeFragment, rnInterpretedText, rnInlineLiteral, rnInlineCode} + if node == nil: + return "" + case node.kind + of rnLeaf, rnSmiley: + result.add node.text + else: + if node.kind in code: result.add "`" + for i in 0 ..< node.sons.len: + if node.kind in {rnInlineCode, rnCodeBlock} and i == 0: + continue # omit language specifier + result.add renderRstToText(node.sons[i]) + if node.kind in code: result.add "`" + +proc treeRepr*(node: PRstNode, indent=0): string = + ## Writes the parsed RST `node` into an AST tree with compact string + ## representation in the format (one line per every sub-node): + ## ``indent - kind - [text|level|order|adType] - anchor (if non-zero)`` + ## (suitable for debugging of RST parsing). + if node == nil: + result.add " ".repeat(indent) & "[nil]\n" + return + result.add " ".repeat(indent) & $node.kind + case node.kind + of rnLeaf, rnSmiley: + result.add (if node.text == "": "" else: " '" & node.text & "'") + of rnEnumList: + result.add " labelFmt=" & node.labelFmt + of rnLineBlockItem: + var txt: string + if node.lineIndent == "\n": txt = " (blank line)" + else: txt = " lineIndent=" & $node.lineIndent.len + result.add txt + of rnAdmonition: + result.add " adType=" & node.adType + of rnHeadline, rnOverline, rnMarkdownHeadline: + result.add " level=" & $node.level + of rnFootnote, rnCitation, rnOptionListItem: + result.add (if node.order == 0: "" else: " order=" & $node.order) + of rnMarkdownBlockQuoteItem: + result.add " quotationDepth=" & $node.quotationDepth + of rnTable, rnGridTable, rnMarkdownTable: + result.add " colCount=" & $node.colCount + of rnTableHeaderCell, rnTableDataCell: + if node.span > 0: + result.add " span=" & $node.span + of rnTableRow: + if node.endsHeader: result.add " endsHeader" + else: + discard + result.add (if node.anchor == "": "" else: " anchor='" & node.anchor & "'") + result.add "\n" + for son in node.sons: + result.add treeRepr(son, indent=indent+2) |