Grammalecte  Check-in [303c416286]

Overview
Comment:[graphspell][fix] tokenizer: new signs
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | graphspell | rg
Files: files | file ages | folders
SHA3-256: 303c416286fcab7c8f0de0a9777e291ed21c429e63d5c5e8f727832040204d05
User & Date: olr on 2018-06-18 20:13:29
Other Links: branch diff | manifest | tags
Context
2018-06-19
06:58
[fr] conversion: regex rules -> graph rules check-in: 086df67518 user: olr tags: fr, rg
2018-06-18
20:13
[graphspell][fix] tokenizer: new signs check-in: 303c416286 user: olr tags: graphspell, rg
20:12
[graphspell] tokenizer: new signs check-in: da0d308818 user: olr tags: graphspell, rg
Changes

Modified graphspell-js/tokenizer.js from [1ab7d18bed] to [5f94dc04ea].

    14     14   const aTkzPatterns = {
    15     15       // All regexps must start with ^.
    16     16       "default":
    17     17           [
    18     18               [/^[   \t]+/, 'SPACE'],
    19     19               [/^\/(?:~|bin|boot|dev|etc|home|lib|mnt|opt|root|sbin|tmp|usr|var|Bureau|Documents|Images|Musique|Public|Téléchargements|Vidéos)(?:\/[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st_.()-]+)*/, 'FOLDERUNIX'],
    20     20               [/^[a-zA-Z]:\\(?:Program Files(?: \(x86\)|)|[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st.()]+)(?:\\[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st_.()-]+)*/, 'FOLDERWIN'],
    21         -            [/^[,.;:!?…«»“”‘’"(){}\[\]/·–—]/, 'SEPARATOR'],
           21  +            [/^[,.;:!?…«»“”‘’"(){}\[\]·–—]/, 'SEPARATOR'],
    22     22               [/^[A-Z][.][A-Z][.](?:[A-Z][.])*/, 'ACRONYM'],
    23     23               [/^(?:https?:\/\/|www[.]|[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st_-]+[@.][a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st_-]{2,}[@.])[a-zA-Z0-9][a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st_.\/?&!%=+*"'@$#-]+/, 'LINK'],
    24     24               [/^[#@][a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st_-]+/, 'TAG'],
    25     25               [/^<[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st]+.*?>|<\/[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st]+ *>/, 'HTML'],
    26     26               [/^\[\/?[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st]+\]/, 'PSEUDOHTML'],
    27     27               [/^&\w+;(?:\w+;|)/, 'HTMLENTITY'],
    28     28               [/^\d\d?h\d\d\b/, 'HOUR'],
................................................................................
    31     31               [/^[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st]+(?:[’'`-][a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st]+)*/, 'WORD']
    32     32           ],
    33     33       "fr":
    34     34           [
    35     35               [/^[   \t]+/, 'SPACE'],
    36     36               [/^\/(?:~|bin|boot|dev|etc|home|lib|mnt|opt|root|sbin|tmp|usr|var|Bureau|Documents|Images|Musique|Public|Téléchargements|Vidéos)(?:\/[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st_.()-]+)*/, 'FOLDERUNIX'],
    37     37               [/^[a-zA-Z]:\\(?:Program Files(?: \(x86\)|)|[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st.()]+)(?:\\[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st_.()-]+)*/, 'FOLDERWIN'],
    38         -            [/^[,.;:!?…«»“”‘’"(){}\[\]/·–—]/, 'SEPARATOR'],
           38  +            [/^[,.;:!?…«»“”‘’"(){}\[\]·–—]/, 'SEPARATOR'],
    39     39               [/^[A-Z][.][A-Z][.](?:[A-Z][.])*/, 'ACRONYM'],
    40     40               [/^(?:https?:\/\/|www[.]|[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st_-]+[@.][a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st_-]{2,}[@.])[a-zA-Z0-9][a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st_.\/?&!%=+*"'@$#-]+/, 'LINK'],
    41     41               [/^[#@][a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st_-]+/, 'TAG'],
    42     42               [/^<[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st]+.*?>|<\/[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st]+ *>/, 'HTML'],
    43     43               [/^\[\/?[a-zA-Zà-öÀ-Ö0-9ø-ÿØ-ßĀ-ʯfi-st]+\]/, 'PSEUDOHTML'],
    44     44               [/^&\w+;(?:\w+;|)/, 'HTMLENTITY'],
    45     45               [/^(?:l|d|n|m|t|s|j|c|ç|lorsqu|puisqu|jusqu|quoiqu|qu)['’`]/i, 'ELPFX'],