diff options
Diffstat (limited to 'lib/pure')
| -rwxr-xr-x | lib/pure/strtabs.nim | 2 | ||||
| -rwxr-xr-x | lib/pure/strutils.nim | 6 | ||||
| -rwxr-xr-x | lib/pure/unidecode/unidecode.nim | 46 |
3 files changed, 31 insertions, 23 deletions
diff --git a/lib/pure/strtabs.nim b/lib/pure/strtabs.nim index 15ff5a079..25220703f 100755 --- a/lib/pure/strtabs.nim +++ b/lib/pure/strtabs.nim @@ -218,7 +218,7 @@ proc `$`*(t: PStringTable): string {.rtl, extern: "nstDollar".} = result.add("}") when isMainModule: - const x = {"k": "v", "11": "22", "565": "67"}.newStringTable + var x = {"k": "v", "11": "22", "565": "67"}.newStringTable assert x["k"] == "v" assert x["11"] == "22" assert x["565"] == "67" diff --git a/lib/pure/strutils.nim b/lib/pure/strutils.nim index 5f91b65b5..33bb5b5f0 100755 --- a/lib/pure/strutils.nim +++ b/lib/pure/strutils.nim @@ -972,9 +972,9 @@ proc c_sprintf(buf, frmt: CString) {.nodecl, importc: "sprintf", varargs, type TFloatFormat* = enum ## the different modes of floating point formating - ffDefault, ## use the shorter floating point notation - ffDecimal, ## use decimal floating point notation - ffScientific ## use scientific notation (using ``e`` character) + ffDefault, ## use the shorter floating point notation + ffDecimal, ## use decimal floating point notation + ffScientific ## use scientific notation (using ``e`` character) proc formatBiggestFloat*(f: BiggestFloat, format: TFloatFormat = ffDefault, precision = 16): string {.noSideEffect, diff --git a/lib/pure/unidecode/unidecode.nim b/lib/pure/unidecode/unidecode.nim index 52f9b6b1a..453e1b289 100755 --- a/lib/pure/unidecode/unidecode.nim +++ b/lib/pure/unidecode/unidecode.nim @@ -1,7 +1,7 @@ # # # Nimrod's Runtime Library -# (c) Copyright 2010 Andreas Rumpf +# (c) Copyright 2011 Andreas Rumpf # # See the file "copying.txt", included in this # distribution, for details about the copyright. @@ -22,23 +22,32 @@ ## strictly one-way transformation. However a human reader will probably ## still be able to guess what original string was meant from the context. ## -## This module needs the data file "unidecode.dat" to work, so it has to be -## shipped with the application! +## This module needs the data file "unidecode.dat" to work: You can either +## ship this file with your application and initialize this module with the +## `loadUnidecodeTable` proc or you can define the ``embedUnidecodeTable`` +## symbol to embed the file as a resource into your application. import unicode -proc loadTranslationTable(filename: string): seq[string] = - newSeq(result, 0xffff) - var i = 0 - for line in lines(filename): - result[i] = line - inc(i) - -var - translationTable: seq[string] +when defined(embedUnidecodeTable): + import strutils -var - datafile* = "unidecode.dat" ## location can be overwritten for deployment + const translationTable = splitLines(slurp"unidecode/unidecode.dat") +else: + # shared is fine for threading: + var translationTable: seq[string] + +proc loadUnidecodeTable*(datafile = "unidecode.dat") = + ## loads the datafile that `unidecode` to work. Unless this module is + ## compiled with the ``embedUnidecodeTable`` symbol defined, this needs + ## to be called by the main thread before any thread can make a call + ## to `unidecode`. + when not defined(embedUnidecodeTable): + newSeq(translationTable, 0xffff) + var i = 0 + for line in lines(datafile): + translationTable[i] = line + inc(i) proc unidecode*(s: string): string = ## Finds the sequence of ASCII characters that is the closest approximation @@ -51,15 +60,14 @@ proc unidecode*(s: string): string = ## ## Results in: "Bei Jing" ## + assert(not isNil(translationTable)) result = "" for r in runes(s): var c = int(r) if c <=% 127: add(result, chr(c)) - elif c <=% 0xffff: - if isNil(translationTable): - translationTable = loadTranslationTable(datafile) - add(result, translationTable[c-128]) + elif c <% translationTable.len: add(result, translationTable[c-128]) -when isMainModule: +when isMainModule: + loadUnidecodeTable("lib/pure/unidecode/unidecode.dat") echo unidecode("Äußerst") |