FIXED: Require automatic abbreviation search to only match abbreviati…

…ons > 1 character long, and only if they are a complete word
fletcher · Nov 18, 2023 · cd8d3b0 · cd8d3b0
1 parent b0b8521
commit cd8d3b0
Show file tree

Hide file tree

Showing 7 changed files with 48 additions and 8 deletions.
diff --git a/src/writer.c b/src/writer.c
@@ -1830,15 +1830,23 @@ void automatic_search_text(mmd_engine * e, token * t, trie * ac) {
 
 	token * tok = t;
 
+	char * source = e->dstr->str;
+
 	if (m) {
 		walker = m->next;
 
 		while (walker) {
-			token_split(tok, walker->start, walker->len, walker->match_type);
-
-			// Advance token to next token
-			while (tok && (tok->start < walker->start + walker->len)) {
-				tok = tok->next;
+			// Only if we match a full word
+			if (
+				(walker->start + walker->len == tok->start + tok->len) ||
+				(char_is_whitespace_or_line_ending_or_punctuation(source[walker->start + walker->len]))
+			) {
+				token_split(tok, walker->start, walker->len, walker->match_type);
+
+				// Advance token to next token
+				while (tok && (tok->start < walker->start + walker->len)) {
+					tok = tok->next;
+				}
 			}
 
 			// Advance to next match (if present)
@@ -1912,13 +1920,19 @@ void identify_global_search_terms(mmd_engine * e, scratch_pad * scratch) {
 	// Add abbreviations to search trie
 	for (int i = 0; i < e->abbreviation_stack->size; ++i) {
 		f = stack_peek_index(e->abbreviation_stack, i);
-		trie_insert(ac, f->label_text, PAIR_BRACKET_ABBREVIATION);
+
+		if (f->label_text && strlen(f->label_text) > 1) {
+			trie_insert(ac, f->label_text, PAIR_BRACKET_ABBREVIATION);
+		}
 	}
 
 	// Add glossary to search trie (without leading '?')
 	for (int i = 0; i < e->glossary_stack->size; ++i) {
 		f = stack_peek_index(e->glossary_stack, i);
-		trie_insert(ac, f->clean_text, PAIR_BRACKET_GLOSSARY);
+
+		if (f->clean_text && strlen(f->clean_text) > 1) {
+			trie_insert(ac, f->clean_text, PAIR_BRACKET_GLOSSARY);
+		}
 	}
 
 	ac_trie_prepare(ac);

diff --git a/tests/MMD6Tests/Abbreviations.fodt b/tests/MMD6Tests/Abbreviations.fodt
@@ -375,6 +375,10 @@ foo</text:p></text:list-item>
 <text:p text:style-name="Standard">foo</text:p></text:list-item>
 
 </text:list>
+
+<text:p text:style-name="Standard">Abbr (Ab) Abbr (A) [>a].</text:p>
+
+<text:p text:style-name="Standard">Abc abc Ab.</text:p>
 </office:text>
 </office:body>
 </office:document>
diff --git a/tests/MMD6Tests/Abbreviations.html b/tests/MMD6Tests/Abbreviations.html
@@ -90,6 +90,10 @@ <h1 id="foobar">foo bar</h1>
 <li><abbr title="FOO">foo</abbr></li>
 </ul>
 
+<p>Abbr (<abbr title="Abbr">Ab</abbr>) Abbr (<abbr title="Abbr">A</abbr>) [>a].</p>
+
+<p>Abc abc <abbr title="Abbr">Ab</abbr>.</p>
+
 <div class="footnotes">
 <hr />
 <ol>

diff --git a/tests/MMD6Tests/Abbreviations.htmlc b/tests/MMD6Tests/Abbreviations.htmlc
@@ -65,6 +65,10 @@ bar</p>
 <li>foo</li>
 </ul>
 
+<p>Ab <a href="Abbr">>A</a> <a href="Abbr">>a</a>.</p>
+
+<p>Abc abc Ab.</p>
+
 <p>[>foo bar]: FOO BAR</p>
 
 <p>[^note]: foo and bar</p>
diff --git a/tests/MMD6Tests/Abbreviations.opml b/tests/MMD6Tests/Abbreviations.opml
@@ -5,7 +5,7 @@
 <outline text="&gt;&gt;Preamble&lt;&lt;" _note="&#10;foo foo&#10;&#10;f.o.o. f.o.o.&#10;&#10;f-o-o f-o-o&#10;&#10;f o o f o o&#10;&#10;fo&apos;o fo&apos;o&#10;&#10;5&#10;&#10;[&gt;bar] [&gt;bar]&#10;&#10;[&gt;b.a.r.] [&gt;b.a.r.]&#10;&#10;[&gt;b-a-r] [&gt;b-a-r]&#10;&#10;[&gt;b a r] [&gt;b a r]&#10;&#10;[&gt;ba&apos;r] [&gt;ba&apos;r]&#10;&#10;10&#10;&#10;foo bar&#10;&#10;foo&#10;bar&#10;&#10;foo  bar&#10;&#10;[&gt;(baz) BAZ]&#10;&#10;&gt; foo bar&#10;&#10;15&#10;&#10;| foo | bar |&#10;| --- | --- |&#10;| foo | bar |&#10;[foo bar]&#10;&#10;"></outline>
 <outline text="foo bar" _note="&#10;"><outline text="foo bar" _note="&#10;"></outline>
 </outline>
-<outline text="foo bar" _note="&#10;![foo bar](http://foobar.net/)&#10;&#10;20&#10;&#10;[^foo and bar]&#10;&#10;[^note]&#10;&#10;* foo&#10;* foo&#10;* foo&#10;&#10;&#10;[&gt;foo]: FOO&#10;[&gt;f.o.o.]: F.O.O.&#10;[&gt;f-o-o]: F-O-O&#10;[&gt;f o o]: F O O&#10;[&gt;fo&apos;o]: FO&apos;O&#10;&#10;[&gt;bar]: BAR&#10;[&gt;b.a.r.]: B.A.R.&#10;[&gt;b-a-r]: B-A-R&#10;[&gt;b a r]: B A R&#10;[&gt;ba&apos;r]: BA&apos;R&#10;&#10;[&gt;foo bar]: FOO BAR&#10;&#10;[^note]: foo and bar&#10;"></outline>
+<outline text="foo bar" _note="&#10;![foo bar](http://foobar.net/)&#10;&#10;20&#10;&#10;[^foo and bar]&#10;&#10;[^note]&#10;&#10;* foo&#10;* foo&#10;* foo&#10;&#10;Ab [&gt;A] [&gt;a].&#10;&#10;Abc abc Ab.&#10;&#10;[&gt;foo]: FOO&#10;[&gt;f.o.o.]: F.O.O.&#10;[&gt;f-o-o]: F-O-O&#10;[&gt;f o o]: F O O&#10;[&gt;fo&apos;o]: FO&apos;O&#10;&#10;[&gt;bar]: BAR&#10;[&gt;b.a.r.]: B.A.R.&#10;[&gt;b-a-r]: B-A-R&#10;[&gt;b a r]: B A R&#10;[&gt;ba&apos;r]: BA&apos;R&#10;&#10;[&gt;foo bar]: FOO BAR&#10;&#10;[^note]: foo and bar&#10;&#10;[&gt;A]: Abbr&#10;[&gt;Ab]: Abbr&#10;"></outline>
 <outline text="&gt;&gt;Metadata&lt;&lt;">
 <outline text="title" _note="Abbreviations"/>
 <outline text="latexconfig" _note="article"/>

diff --git a/tests/MMD6Tests/Abbreviations.tex b/tests/MMD6Tests/Abbreviations.tex
@@ -22,6 +22,10 @@
 
 \newacronym{foo bar}{foo bar}{FOO BAR}
 
+\newacronym{A}{A}{Abbr}
+
+\newacronym{Ab}{Ab}{Abbr}
+
 \input{mmd6-article-begin}
 
 \gls{foo} \gls{foo}
@@ -111,5 +115,9 @@ \part{foo bar}
 
 \end{itemize}
 
+\gls{Ab} \gls{A} {[>a]}.
+
+Abc abc \gls{Ab}.
+
 \input{mmd6-article-footer}
 \end{document}
diff --git a/tests/MMD6Tests/Abbreviations.text b/tests/MMD6Tests/Abbreviations.text
@@ -62,6 +62,9 @@ foo bar
 * foo
 * foo
 
+Ab [>A] [>a].
+
+Abc abc Ab.
 
 [>foo]: FOO
 [>f.o.o.]: F.O.O.
@@ -78,3 +81,6 @@ foo bar
 [>foo bar]: FOO BAR
 
 [^note]: foo and bar
+
+[>A]: Abbr
+[>Ab]: Abbr