Is anyone successfully using the custom regular expression feature (when generating the description automatically from the teaser)? I can't seem to get it to do anything. I'm still learning regex, but even if I simplify it down to just one word (when testing it out) I still can't get it to strip that word...
| Comment | File | Size | Author |
|---|---|---|---|
| #15 | nodewords-n983550-15.patch | 4.73 KB | damienmckenna |
| #14 | nodewords-n983550-14.patch | 4.79 KB | damienmckenna |
| #7 | nodewords_regex_fix.patch | 5.9 KB | quicksketch |
| #5 | nodewords_regex_fix.patch | 5.88 KB | quicksketch |
| #4 | nodewords_regex_fix.patch | 5.69 KB | quicksketch |
Comments
Comment #1
dave reidComment #2
Jiri Volf commentedDoes not work for me either. Tried the img_assist.module checkbox to get rid of image assist code in automatically generated descriptions, but it did not work, so tried to put
s/[.*]//
s/[.*]//gi
and other variations into the regex field, but did not work either..
drupal 6.19, modules: Better Formats, Wysiwyg, tinymce, Views, Nodewords, Image assist
btw, also failed to set up special tags for other pages, it always reverts to default values...
Comment #3
WildBill commentedThe img_assist.module checkbox works for me, stripping out all tags created by the Image Assist module. But I'd also like to strip out tags created by the Zemanta module (super-handy module, by the way), but I can't get the custom regex box to do anything.
Using the same setup as vlx above, but instead of tinymce, I'm using CKEditor (in conjunction with WYSIWYG).
Comment #4
quicksketchThe regex already included the leading and trailing slashes (confusing I know). I think this would best be served by putting #field_prefix and #field_suffix on the field, then you'd know they're there already.
But in addition to that, the regexs aren't even run due to a typo in the code "$regexp" vs. "$regex". Finally, the regexs are only run if generating from a teaser, not if generating from a body or the option for "teaser if body is empty". It really doesn't make any sense. This patch cleans up the code for consistent cleanup of the description text no matter its source as well as running the regexs properly.
Comment #5
quicksketchMy last patch didn't check for an empty teaser correctly. In the future, I think we could actually safely ditch the "Generate from body if teaser is empty" option. Obviously (90% use-case) if you selected the teaser option you'd want this as the default anyway.
Comment #7
quicksketchApparently Git's a/b prefixes are now necessary for testing.
Comment #9
quicksketchI'll figure out this testbot yet. :P
Comment #10
quicksketch#7: nodewords_regex_fix.patch queued for re-testing.
Comment #11
dave reidComment #12
dave reidComment #14
damienmckennaRerolled. Needs some thorough testing before I'll commit it.
Comment #15
damienmckennaRerolled. It also fixes a logic bug when the teaser was not empty but the auto-generate option was set to "teaser or body".
Comment #16
damienmckennaCommitted.