Further cleanup of ts_headline code.

tglsfdc · tglsfdc · commit 2e0e409e3cba · 2020-04-09T15:38:43.000-04:00
Suppress a probably-meaningless uninitialized-variable warning (induced by my previous patch, I'm sorry to say). Improve mark_hl_fragments()'s test for overlapping cover strings: it failed to consider the possibility that the current string is strictly within another one. That's unlikely given the preceding splitting into MaxWords fragments, but I don't think it's impossible. Discussion: https://postgr.es/m/16345-2e0cf5cddbdcd3b4@postgresql.org
diff --git a/src/backend/tsearch/wparser_def.c b/src/backend/tsearch/wparser_def.c
@@ -2337,22 +2337,24 @@ mark_hl_fragments(HeadlineParsedText *prs, TSQuery query, bool highlightall,
 			/* Mark the chosen fragments (covers) */
 			mark_fragment(prs, highlightall, startpos, endpos);
 			num_f++;
-			/* exclude overlapping covers */
+			/* Exclude covers overlapping this one from future consideration */
 			for (i = 0; i < numcovers; i++)
 			{
 				if (i != minI &&
-					((covers[i].startpos >= covers[minI].startpos &&
-					  covers[i].startpos <= covers[minI].endpos) ||
-					 (covers[i].endpos >= covers[minI].startpos &&
-					  covers[i].endpos <= covers[minI].endpos)))
+					((covers[i].startpos >= startpos &&
+					  covers[i].startpos <= endpos) ||
+					 (covers[i].endpos >= startpos &&
+					  covers[i].endpos <= endpos) ||
+					 (covers[i].startpos < startpos &&
+					  covers[i].endpos > endpos)))
 					covers[i].excluded = true;
 			}
 		}
 		else
-			break;
+			break;				/* no selectable covers remain */
 	}
 
-	/* show at least min_words if we have not marked anything */
+	/* show the first min_words words if we have not marked anything */
 	if (num_f <= 0)
 	{
 		startpos = endpos = curlen = 0;
@@ -2510,6 +2512,7 @@ mark_hl_words(HeadlineParsedText *prs, TSQuery query, bool highlightall,
 		if (bestlen < 0)
 		{
 			curlen = 0;
+			pose = 0;
 			for (i = 0; i < prs->curwords && curlen < min_words; i++)
 			{
 				if (!NONWORDTOKEN(prs->words[i].type))

Original file line number	Diff line number	Diff line change
`@@ -2337,22 +2337,24 @@ mark_hl_fragments(HeadlineParsedText *prs, TSQuery query, bool highlightall,`
`2337`	`2337`	`/* Mark the chosen fragments (covers) */`
`2338`	`2338`	`mark_fragment(prs, highlightall, startpos, endpos);`
`2339`	`2339`	`num_f++;`
`2340`		`- /* exclude overlapping covers */`
	`2340`	`+ /* Exclude covers overlapping this one from future consideration */`
`2341`	`2341`	`for (i = 0; i < numcovers; i++)`
`2342`	`2342`	`{`
`2343`	`2343`	`if (i != minI &&`
`2344`		`- ((covers[i].startpos >= covers[minI].startpos &&`
`2345`		`- covers[i].startpos <= covers[minI].endpos) \|\|`
`2346`		`- (covers[i].endpos >= covers[minI].startpos &&`
`2347`		`- covers[i].endpos <= covers[minI].endpos)))`
	`2344`	`+ ((covers[i].startpos >= startpos &&`
	`2345`	`+ covers[i].startpos <= endpos) \|\|`
	`2346`	`+ (covers[i].endpos >= startpos &&`
	`2347`	`+ covers[i].endpos <= endpos) \|\|`
	`2348`	`+ (covers[i].startpos < startpos &&`
	`2349`	`+ covers[i].endpos > endpos)))`
`2348`	`2350`	`covers[i].excluded = true;`
`2349`	`2351`	`}`
`2350`	`2352`	`}`
`2351`	`2353`	`else`
`2352`		`- break;`
	`2354`	`+ break; /* no selectable covers remain */`
`2353`	`2355`	`}`
`2354`	`2356`
`2355`		`- /* show at least min_words if we have not marked anything */`
	`2357`	`+ /* show the first min_words words if we have not marked anything */`
`2356`	`2358`	`if (num_f <= 0)`
`2357`	`2359`	`{`
`2358`	`2360`	`startpos = endpos = curlen = 0;`
`@@ -2510,6 +2512,7 @@ mark_hl_words(HeadlineParsedText *prs, TSQuery query, bool highlightall,`
`2510`	`2512`	`if (bestlen < 0)`
`2511`	`2513`	`{`
`2512`	`2514`	`curlen = 0;`
	`2515`	`+ pose = 0;`
`2513`	`2516`	`for (i = 0; i < prs->curwords && curlen < min_words; i++)`
`2514`	`2517`	`{`
`2515`	`2518`	`if (!NONWORDTOKEN(prs->words[i].type))`