From d0180007366dd9e1e4b1e7aec249998135fbaa2d Mon Sep 17 00:00:00 2001 From: Jan Nieuwenhuizen Date: Fri, 6 Mar 2009 11:31:28 +0100 Subject: [PATCH] 2009-01-29 Jan Nieuwenhuizen * src/yelp-info-parser.c (info_body_text): Also substitute Notes, greatly simplifying previous: (info_process_text_notes): Remove. (parse_tree_level): (yelp_info_parse_menu): Update callers. --- ChangeLog | 8 + src/yelp-info-parser.c | 347 +++++++++++++----------------------------------- 2 files changed, 100 insertions(+), 255 deletions(-) diff --git a/ChangeLog b/ChangeLog index c2061e8..9078892 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,11 @@ +2009-01-29 Jan Nieuwenhuizen + + * src/yelp-info-parser.c (info_body_text): Also substitute + Notes, greatly simplifying previous: + (info_process_text_notes): Remove. + (parse_tree_level): + (yelp_info_parse_menu): Update callers. + 2009-01-18 Jan Nieuwenhuizen * src/yelp-info-parser.c (info_image_get_attributes): diff --git a/src/yelp-info-parser.c b/src/yelp-info-parser.c index c5d38ca..8b623e3 100644 --- a/src/yelp-info-parser.c +++ b/src/yelp-info-parser.c @@ -42,8 +42,7 @@ GtkTreeIter * find_real_sibling (GtkTreeModel *model, GtkTreeIter *comp); xmlNodePtr yelp_info_parse_menu (GtkTreeStore *tree, xmlNodePtr *node, - gchar *page_content, - gboolean notes); + gchar *page_content); gboolean get_menuoptions (gchar *line, gchar **title, gchar **ref, @@ -56,9 +55,6 @@ gboolean resolve_frag_id (GtkTreeModel *model, void fix_tag_table (gchar *offset, gpointer page, TagTableFix *a); -void info_process_text_notes (xmlNodePtr *node, - gchar *content, - GtkTreeStore *tree); static GHashTable * @@ -141,39 +137,108 @@ info_insert_image (xmlNodePtr parent, GMatchInfo *match_info) return parent; } +static xmlNodePtr +info_insert_note (xmlNodePtr parent, GMatchInfo *match_info) +{ + gchar* name = g_match_info_fetch (match_info, 1); + gchar* file = g_match_info_fetch (match_info, 2); + gchar* node = g_match_info_fetch (match_info, 3); + xmlNodePtr a = xmlNewTextChild (parent, NULL, BAD_CAST "a", BAD_CAST (name ? name : (node ? node : "Top"))); + gchar* href = g_strconcat ("info:", (file ? file : ""), "#", (node ? node : "Top"), NULL); + xmlNewProp (a, BAD_CAST "href", BAD_CAST href); + g_free (name); + g_free (file); + g_free (node); + g_free (href); + return parent; +} + +static GMatchInfo* +regex_string_get_match_info (GRegex* regex, gchar const* string) +{ + GMatchInfo *match_info; + g_regex_match (regex, string, 0, &match_info); + g_regex_unref (regex); + return match_info; +} + +static GMatchInfo* +string_get_image_match_info (gchar const* string) +{ + return regex_string_get_match_info (g_regex_new ("(" INFO_C_IMAGE_TAG_OPEN_RE "((?:[^" INFO_TAG_1 "]|[^" INFO_C_TAG_0 "]+" INFO_TAG_1 ")*)" INFO_C_TAG_CLOSE_RE ")", 0, 0, NULL), string); +} + +static GMatchInfo* +string_get_note_match_info (gchar const* string) +{ + return regex_string_get_match_info (g_regex_new ("[*][Nn]ote\\s(?:([^:]+):)?\\s*(?:\\(([^:()]+)(?:(?:(?:[.]info)\\))|\\)))?\\s*([^():.,]+)(?:::)?", 0, 0, NULL), string); +} + /* Convert body text CONTENT to xml nodes, processing info image tags - when found. IWBN add a regex match for *Note: here and call the - *Note ==> logic of info_process_text_notes from here. + and text notes when found. */ static xmlNodePtr -info_body_text (xmlNodePtr parent, xmlNsPtr ns, gchar const *name, gchar const *content) +info_body_text (xmlNodePtr parent, gchar const *name, gchar *content) { - if (!strstr (content, INFO_C_IMAGE_TAG_OPEN)) - return xmlNewTextChild (parent, ns, BAD_CAST name, BAD_CAST content); + if (!strstr (content, INFO_C_IMAGE_TAG_OPEN) + && !strstr (content, "*Note ") + && !strstr (content, "*note ")) + return xmlNewTextChild (parent, NULL, BAD_CAST name, BAD_CAST content); - gint content_len = strlen (content); gint pos = 0; - GRegex *regex = g_regex_new ("(" INFO_C_IMAGE_TAG_OPEN_RE "((?:[^" INFO_TAG_1 "]|[^" INFO_C_TAG_0 "]+" INFO_TAG_1 ")*)" INFO_C_TAG_CLOSE_RE ")", 0, 0, NULL); - GMatchInfo *match_info; - g_regex_match (regex, content, 0, &match_info); - while (g_match_info_matches (match_info)) + gint content_len = strlen (content); + GMatchInfo *image_match_info = string_get_image_match_info (content); + GMatchInfo *note_match_info = string_get_note_match_info (content); + + while (g_match_info_matches (image_match_info) + || g_match_info_matches (note_match_info)) { gint image_start; gint image_end; - gboolean image_found = g_match_info_fetch_pos (match_info, 0, + gboolean image_found = g_match_info_fetch_pos (image_match_info, 0, &image_start, &image_end); - gchar *before = g_strndup (&content[pos], image_start - pos); - pos = image_end + 1; + gint note_start; + gint note_end; + gboolean note_found = g_match_info_fetch_pos (note_match_info, 0, + ¬e_start, ¬e_end); + gint start; + gint end; + if (image_found + && (!note_found + || (image_start < note_start))) + { + note_found = FALSE; + start = image_start; + end = image_end; + } + else // note_found + { + image_found = FALSE; + start = note_start; + end = note_end; + } + + gchar *before = g_strndup (&content[pos], start - pos); + pos = end + 1; xmlNewTextChild (parent, NULL, BAD_CAST "para1", BAD_CAST (before)); g_free (before); if (image_found) - info_insert_image (parent, match_info); - g_match_info_next (match_info, NULL); + { + info_insert_image (parent, image_match_info); + g_match_info_next (image_match_info, NULL); + } + else // (note_found) + { + info_insert_note (parent, note_match_info); + g_match_info_next (note_match_info, NULL); + } } gchar *after = g_strndup (&content[pos], content_len - pos); xmlNewTextChild (parent, NULL, BAD_CAST "para1", BAD_CAST (after)); g_free (after); + g_match_info_free (image_match_info); + g_match_info_free (note_match_info); return 0; } @@ -819,7 +884,6 @@ parse_tree_level (GtkTreeStore *tree, xmlNodePtr *node, GtkTreeIter iter) char *page_no = NULL; char *page_name = NULL; char *page_content = NULL; - gboolean notes = FALSE; debug_print (DB_DEBUG, "Decended\n"); do @@ -830,23 +894,13 @@ parse_tree_level (GtkTreeStore *tree, xmlNodePtr *node, GtkTreeIter iter) COLUMN_PAGE_CONTENT, &page_content, -1); debug_print (DB_DEBUG, "Got Section: %s\n", page_name); - if (strstr (page_content, "*Note") || - strstr (page_content, "*note")) { - notes = TRUE; - } if (strstr (page_content, "* Menu:")) { - newnode = yelp_info_parse_menu (tree, node, page_content, notes); + newnode = yelp_info_parse_menu (tree, node, page_content); } else { newnode = xmlNewTextChild (*node, NULL, BAD_CAST "Section", NULL); - if (!notes) - info_body_text (newnode, NULL, "para", page_content); - - else { - /* Handle notes here */ - info_process_text_notes (&newnode, page_content, tree); - } + info_body_text (newnode, "para1", page_content); } /* if we free the page content, now it's in the XML, we can * save some memory */ @@ -899,6 +953,9 @@ yelp_info_parser_parse_tree (GtkTreeStore *tree) g_print ("XML follows:\n%s\n", xmlbuf); */ + if (getenv ("YELP_INFO_DUMP_DOC")) + xmlDocDump (stderr, doc); + return doc; } @@ -981,7 +1038,7 @@ get_menuoptions (gchar *line, gchar **title, gchar **ref, gchar **desc, xmlNodePtr yelp_info_parse_menu (GtkTreeStore *tree, xmlNodePtr *node, - gchar *page_content, gboolean notes) + gchar *page_content) { gchar **split; gchar **menuitems; @@ -996,12 +1053,7 @@ yelp_info_parse_menu (GtkTreeStore *tree, xmlNodePtr *node, tmp = g_strconcat (split[0], "\n* Menu:", NULL); - if (!notes) - xmlNewTextChild (newnode, NULL, - BAD_CAST "para", BAD_CAST tmp); - else { - info_process_text_notes (&newnode, tmp, tree); - } + info_body_text (newnode, "para1", tmp); g_free (tmp); menuitems = g_strsplit (split[1], "\n", -1); @@ -1078,218 +1130,3 @@ yelp_info_parse_menu (GtkTreeStore *tree, xmlNodePtr *node, return newnode; } - -void -info_process_text_notes (xmlNodePtr *node, gchar *content, GtkTreeStore *tree) -{ - gchar **notes; - gchar **current; - xmlNodePtr holder; - xmlNodePtr ref1; - gboolean first = TRUE; - - notes = g_strsplit (content, "*Note", -1); - holder = xmlNewChild (*node, NULL, BAD_CAST "noteholder", NULL); - - for (current = notes; *current != NULL; current++) { - /* Since the notes can be either *Note or *note, we handle the second - * variety here - */ - gchar **subnotes; - gchar **current_real; - - subnotes = g_strsplit (*current, "*note", -1); - for (current_real = subnotes; *current_real != NULL; current_real++) { - gchar *url, **urls, **ulink; - gchar *append; - gchar *alt_append, *alt_append1; - gchar *link_text; - gchar *href = NULL; - gchar *break_point = NULL; - gboolean broken = FALSE; - if (first) { - /* The first node is special. It doesn't have a note ref at the - * start, so we can just add it and forget about it. - */ - first = FALSE; - info_body_text (holder, NULL, "para1", (*current_real)); - continue; - } - /* If we got to here, we now gotta parse the note reference */ - - if (*current_real[0] == '_') { - /* Special type of note that isn't really a note, but pretends - * it is - */ - info_body_text (holder, NULL, "para1", - g_strconcat ("*Note", *current_real, NULL)); - continue; - } - append = strchr (*current_real, ':'); - if (!append) { - info_body_text (holder, NULL, "para1", *current_real); - continue; - } - append++; - alt_append = append; - alt_append1 = alt_append; - append = strchr (append, ':'); - alt_append = strchr (alt_append, '.'); - if (alt_append && g_str_has_prefix (alt_append, ".info")) { - broken = TRUE; - alt_append++; - alt_append = strchr (alt_append, '.'); - } - alt_append1 = strchr (alt_append1, ','); - if (!append && !alt_append && !alt_append1) { - info_body_text (holder, NULL, "para1", *current_real); - continue; - } - if (!append || alt_append || alt_append1) { - if (!append) { - if (alt_append) append = alt_append; - else append = alt_append1; - } - if ((alt_append && alt_append < append)) - append = alt_append; - if (alt_append1 && alt_append1 < append) - append = alt_append1; - } - append++; - url = g_strndup (*current_real, append - (*current_real)); - - /* By now, we got 2 things. First, is append which is the (hopefully) - * non-link text. Second, we got a url. - * The url can be in several forms: - * 1. linkend:: - * 2. linkend:(infofile)Linkend. - * 3. Title: Linkend. - * 4. Title: Linkend, (pretty sure this is just broken) - * 5. Title: (infofile.info)Linkend. - * All possibilities should have been picked up. - * Here: - * Clean up the split. Should be left with a real url and - * a list of fragments that should be linked - * Also goes through and removes extra spaces, leaving only one - * space in place of many - */ - urls = g_strsplit (url, "\n", -1); - break_point = strchr (url, '\n'); - while (break_point) { - *break_point = ' '; - break_point = strchr (++break_point, '\n'); - } - break_point = strchr (url, ' '); - while (break_point) { - if (*(break_point+1) == ' ') { - /* Massive space. Fix. */ - gchar *next = break_point; - gchar *url_copy; - while (*next == ' ') - next++; - next--; - url_copy = g_strndup (url, break_point-url); - g_free (url); - url = g_strconcat (url_copy, next, NULL); - break_point = strchr (url, ' '); - g_free (url_copy); - } else { - break_point++; - break_point = strchr (break_point, ' '); - } - } - if (url[strlen(url)-1] == '.') { /* The 2nd or 3rd sort of link */ - gchar *stop = NULL; - gchar *lurl = NULL; - gchar *zloc = NULL; - stop = strchr (url, ':'); - lurl = strchr (stop, '('); - if (!lurl) { /* 3rd type of link */ - gchar *link; - gint length; - stop++; - link = g_strdup (stop); - link = g_strstrip (link); - length = strlen (link) - 1; - link[length] = '\0'; - href = g_strconcat ("#", link, NULL); - link[length] = 'a'; - g_free (link); - - - } else { /* 2nd type of link. Easy. Provided .info is neglected ;) */ - if (broken) { - gchar *new_url; - gchar *info; - gchar *stripped; - - new_url = g_strdup (lurl); - info = strstr (new_url, ".info)"); - stripped = g_strndup (new_url, info-new_url); - info +=5; - lurl = g_strconcat (stripped, info, NULL); - g_free (stripped); - g_free (new_url); - } - zloc = &(lurl[strlen(lurl)-1]); - *zloc = '\0'; - href = g_strconcat ("info:", lurl, NULL); - *zloc = 'a'; - } - } else { /* First kind of link */ - gchar *tmp1; - gchar *frag; - - tmp1 = strchr (url, ':'); - if (!tmp1) - frag = g_strdup (url); - else - frag = g_strndup (url, tmp1 - url); - g_strstrip (frag); - gtk_tree_model_foreach (GTK_TREE_MODEL (tree), resolve_frag_id, &frag); - href = g_strconcat ("#", frag, NULL); - g_free (frag); - } - for (ulink = urls; *ulink != NULL; ulink++) { - if (ulink == urls) - link_text = g_strconcat ("*Note", *ulink, NULL); - else { - gchar *spacing = *ulink; - gchar *tmp; - gint count = 0; - while (*spacing == ' ') { - spacing++; - count++; - } - if (spacing != *ulink) { - if (count > 1) - spacing-=2; - tmp = g_strndup (*ulink, spacing-*ulink); - if (count > 1) - spacing+=2; - xmlNewTextChild (holder, NULL, BAD_CAST "spacing", - BAD_CAST tmp); - g_free (tmp); - link_text = g_strdup (spacing); - } else { - link_text = g_strdup (*ulink); - } - } - ref1 = xmlNewTextChild (holder, NULL, BAD_CAST "a", - BAD_CAST link_text); - if (*(ulink+1) != NULL) - info_body_text (holder, NULL, "para", ""); - - g_free (link_text); - xmlNewProp (ref1, BAD_CAST "href", BAD_CAST href); - } - g_strfreev (urls); - /* Finally, we can add the text as required */ - info_body_text (holder, NULL, "para1", append); - g_free (url); - g_free (href); - } - g_strfreev (subnotes); - } - g_strfreev (notes); -} -- 1.6.0.rc1.49.g98a8