Commit graph

33 commits

Author SHA1 Message Date
acli
de8d0930ee Minor factoring of construction of warning messages. 2004-02-16 22:50:34 +00:00
acli
2a9be2b2e6 Don't bother warning about TMPL_VAR if the key is onclick, onblur, etc.
We don't know how to warn/what to suggest, & that will only confuse people
2004-02-14 09:50:11 +00:00
acli
1d45c47c02 Fix spurious warnings if attribute is in the form foo="bar"</TMPL_IF> 2004-02-14 09:41:28 +00:00
acli
f7b649f41b Make a reasonable suggestion for ESCAPE= if we warn about lack of it 2004-02-14 09:33:09 +00:00
acli
3fd0a52e0a Fixed spurious warning about unescaped < inside cdata 2004-02-14 09:23:34 +00:00
acli
050e1995d9 Minor change to make the "closed start tag" warning more understandable 2004-02-14 09:10:20 +00:00
acli
ce2189ef37 Don't complain about strange attribute syntax if what we see is a
reasonable templating control flow directive (if, else, unless).
2004-02-14 08:49:21 +00:00
acli
524a76f1b3 Have to make it know what "closed start tag" notation is; other it spews
out more than a screenful or text for an "unknown token" when such notation
is seen
2004-02-14 08:03:02 +00:00
acli
1b95b1698e Fixed problem recognizing tags in the form of <foo ... bar=<!-- TMPL_VAR ... >> 2004-02-14 07:49:37 +00:00
acli
2afa50bdda Don't extract TMPL_VAR's as if they were translatable 2004-02-14 07:13:09 +00:00
acli
10bec90dea Don't complain about </TMPL_IF> or </TMPL_LOOP> being strange attribute
syntax; they are fine.

The way TMPL_VAR is warned probably makes more sense now.
2004-02-14 07:07:36 +00:00
acli
0e2ff58b66 This should be still more correct regarding when to warn about TMPL_VAR
in attributes
2004-02-14 06:16:36 +00:00
acli
16992ec3f9 More correct version of previous change 2004-02-14 05:50:55 +00:00
acli
93740ec7ba Make sure that if an attribute contains < or >, a warning is given; these
warnings aren't pedantic because (1) if it's a templating directive, it
might expand into something containing a real < and/or >, and (2) if it
contains >, the browser will close the current tag, and (3) if it contains
< and the browser knows what "SGML closed start tags" are (e.g., Mozilla),
the browser will also close the current tag.
2004-02-14 05:46:38 +00:00
acli
a2f07d16f0 Hmm. I didn't know there can be whitespace before and/or after = in
attributes...
2004-02-14 05:35:04 +00:00
acli
a10bb7062a Handle leading or trailing &nbsp;'s as leading or trailing spaces.
Make sure they are all trimmed off.

$re_end_entity is now used (by the above); there are &nbsp's (no ;)
in our templates.
2004-02-13 03:49:26 +00:00
acli
b87b492773 The fixed search.marc/search.tmpl (nothing between <textarea></textarea>)
caused an eof token to be incorrectly generated by next_token(). This
is now fixed.
2004-02-13 02:42:06 +00:00
acli
5af84e39db Don't mindlessly spew out 40 lines of text in a warning message if we get
too confused.  Just say we are too confused.
2004-02-13 01:48:58 +00:00
acli
412847fe93 This way of reporting line numbers should make more sense,
esp. for pathetic cases like search.marc/search.tmpl
(missing closing " for an attribute)
2004-02-13 01:27:03 +00:00
acli
1c3cb74b82 Display something useful if the user doesn't specify -f 2004-02-13 01:20:03 +00:00
acli
1f128d7381 Don't issue warnings for unquoted attributes containing [^-\.a-zA-Z0-9]
unless --pedantic-warnings is given. These don't seem to cause any trouble,
even in Mozilla's standards compliant mode.
2004-02-13 01:14:18 +00:00
acli
1addd34bb1 Paul's problem #1 is now fixed: Bug in regular expression $re_directive.
Put my "grander plan" :-) in the comments
2004-02-13 01:03:18 +00:00
acli
250d1fcefc Don't extract purely-numeric strings like "1" either 2004-02-13 00:48:52 +00:00
acli
a49634cb34 Seems like I wasn't careful enough recognizing unknown tokens. Incomplete
tags like "<b foo" at the end of the file seems to be discarded silently by
Mozilla, even in quirks mode. We now display a warning for these (in case
these ever come up by accident).
2004-02-13 00:42:52 +00:00
acli
bed495ed3e Minor wording rewrite in warning 2004-02-12 18:25:43 +00:00
acli
906bfbc3d6 Warn about unquoted attribute values containing [^-\.a-zA-Z0-9] 2004-02-12 18:19:17 +00:00
acli
21403fd5cc extract_attributes now knows what XML-style self-closing tags are 2004-02-12 17:58:24 +00:00
acli
0a9cd4aba1 Warn about <<Prev 2004-02-12 17:44:59 +00:00
acli
33a4d5705a It now knows what << Prev is, but it may still be confused by other kinds
of tag lookalikes
2004-02-12 17:38:42 +00:00
acli
5affdbf4e7 Handle attributes which are TMPL_VAR's better. It was choking on
systempreferences.tmpl.
2004-02-12 09:38:20 +00:00
acli
2f928c4e75 Don't extract strings in hidden values 2004-02-12 09:26:54 +00:00
acli
0be46ba475 This should be good enough to replace text-extract.pl, but some real
testing is needed.
2004-02-12 09:02:39 +00:00
acli
7d244a0b70 This is an experimental filter, based on simple scanning, that *should*
(ultimately) work better than the standard filter based on real parsing
of the .tmpl files.
2004-02-12 08:55:14 +00:00