X-Git-Url: http://sjero.net/git/?p=wget;a=blobdiff_plain;f=src%2Fhtml-url.c;h=09962eddc9286a0f6a6569c697660caf9f2e96a8;hp=abaa2f8e59bcdc18007caa60c4c55fbc75f832e8;hb=ae1d264fcc190f9c74cb490aa6da0240b0b77b1e;hpb=a9c3c58c9fb22e71e0878d4da1d3de0cd9e36445
diff --git a/src/html-url.c b/src/html-url.c
index abaa2f8e..09962edd 100644
--- a/src/html-url.c
+++ b/src/html-url.c
@@ -643,6 +643,7 @@ get_urls_html (const char *file, const char *url, int *meta_disallow_follow)
{
struct file_memory *fm;
struct map_context ctx;
+ int flags;
/* Load the file. */
fm = read_file (file);
@@ -663,8 +664,16 @@ get_urls_html (const char *file, const char *url, int *meta_disallow_follow)
if (!interesting_tags)
init_interesting ();
- map_html_tags (fm->content, fm->length, interesting_tags,
- interesting_attributes, collect_tags_mapper, &ctx);
+ /* Specify MHT_TRIM_VALUES because of buggy HTML generators that
+ generate instead of (Netscape
+ ignores spaces as well.) If you really mean space, use &32; or
+ %20. */
+ flags = MHT_TRIM_VALUES;
+ if (opt.strict_comments)
+ flags |= MHT_STRICT_COMMENTS;
+
+ map_html_tags (fm->content, fm->length, collect_tags_mapper, &ctx, flags,
+ interesting_tags, interesting_attributes);
DEBUGP (("no-follow in %s: %d\n", file, ctx.nofollow));
if (meta_disallow_follow)