mirrors/pdf.js
1
0
Fork 0
mirror of https://github.com/mozilla/pdf.js.git synced 2025-04-29 15:47:57 +02:00
Code Activity

Merge pull request #5579 from jazzy-em/master

Browse source 
Added multiple term search functionality (with default phrase search)
This commit is contained in:
Yury Delendik 2016-05-27 14:23:26 -05:00
commit 5aefce6ba6
7 changed files with 154 additions and 16 deletions
Download patch file Download diff file Expand all files Collapse all files

124
web/pdf_find_controller.js
View file

@ -78,6 +78,7 @@ var PDFFindController = (function PDFFindControllerClosure() {
this.active = false; // If active, find results will be highlighted.
this.pageContents = []; // Stores the text for each page.
this.pageMatches = [];
this.pageMatchesLength = null;
this.matchCount = 0;
this.selected = { // Currently selected match.
pageIdx: -1,
@ -104,10 +105,114 @@ var PDFFindController = (function PDFFindControllerClosure() {
});
},
// Helper for multiple search - fills matchesWithLength array
// and takes into account cases when one search term
// include another search term (for example, "tamed tame" or "this is").
// Looking for intersecting terms in the 'matches' and
// leave elements with a longer match-length.
_prepareMatches: function PDFFindController_prepareMatches(
matchesWithLength, matches, matchesLength) {
function isSubTerm(matchesWithLength, currentIndex) {
var currentElem, prevElem, nextElem;
currentElem = matchesWithLength[currentIndex];
nextElem = matchesWithLength[currentIndex + 1];
// checking for cases like "TAMEd TAME"
if (currentIndex < matchesWithLength.length - 1 &&
currentElem.match === nextElem.match) {
currentElem.skipped = true;
return true;
}
// checking for cases like "thIS IS"
for (var i = currentIndex - 1; i >= 0; i--) {
prevElem = matchesWithLength[i];
if (prevElem.skipped) {
continue;
}
if (prevElem.match + prevElem.matchLength < currentElem.match) {
break;
}
if (prevElem.match + prevElem.matchLength >=
currentElem.match + currentElem.matchLength) {
currentElem.skipped = true;
return true;
}
}
return false;
}
var i, len;
// Sorting array of objects { match: <match>, matchLength: <matchLength> }
// in increasing index first and then the lengths.
matchesWithLength.sort(function(a, b) {
return a.match === b.match ?
a.matchLength - b.matchLength : a.match - b.match;
});
for (i = 0, len = matchesWithLength.length; i < len; i++) {
if (isSubTerm(matchesWithLength, i)) {
continue;
}
matches.push(matchesWithLength[i].match);
matchesLength.push(matchesWithLength[i].matchLength);
}
},
calcFindPhraseMatch: function PDFFindController_calcFindPhraseMatch(
query, pageIndex, pageContent) {
var matches = [];
var queryLen = query.length;
var matchIdx = -queryLen;
while (true) {
matchIdx = pageContent.indexOf(query, matchIdx + queryLen);
if (matchIdx === -1) {
break;
}
matches.push(matchIdx);
}
this.pageMatches[pageIndex] = matches;
},
calcFindWordMatch: function PDFFindController_calcFindWordMatch(
query, pageIndex, pageContent) {
var matchesWithLength = [];
// Divide the query into pieces and search for text on each piece.
var queryArray = query.match(/\S+/g);
var subquery, subqueryLen, matchIdx;
for (var i = 0, len = queryArray.length; i < len; i++) {
subquery = queryArray[i];
subqueryLen = subquery.length;
matchIdx = -subqueryLen;
while (true) {
matchIdx = pageContent.indexOf(subquery, matchIdx + subqueryLen);
if (matchIdx === -1) {
break;
}
// Other searches do not, so we store the length.
matchesWithLength.push({
match: matchIdx,
matchLength: subqueryLen,
skipped: false
});
}
}
// Prepare arrays for store the matches.
if (!this.pageMatchesLength) {
this.pageMatchesLength = [];
}
this.pageMatchesLength[pageIndex] = [];
this.pageMatches[pageIndex] = [];
// Sort matchesWithLength, clean up intersecting terms
// and put the result into the two arrays.
this._prepareMatches(matchesWithLength, this.pageMatches[pageIndex],
this.pageMatchesLength[pageIndex]);
},
calcFindMatch: function PDFFindController_calcFindMatch(pageIndex) {
var pageContent = this.normalize(this.pageContents[pageIndex]);
var query = this.normalize(this.state.query);
var caseSensitive = this.state.caseSensitive;
var phraseSearch = this.state.phraseSearch;
var queryLen = query.length;
if (queryLen === 0) {
@ -120,16 +225,12 @@ var PDFFindController = (function PDFFindControllerClosure() {
query = query.toLowerCase();
}
var matches = [];
var matchIdx = -queryLen;
while (true) {
matchIdx = pageContent.indexOf(query, matchIdx + queryLen);
if (matchIdx === -1) {
break;
}
matches.push(matchIdx);
if (phraseSearch) {
this.calcFindPhraseMatch(query, pageIndex, pageContent);
} else {
this.calcFindWordMatch(query, pageIndex, pageContent);
}
this.pageMatches[pageIndex] = matches;
this.updatePage(pageIndex);
if (this.resumePageIdx === pageIndex) {
this.resumePageIdx = null;
@ -137,8 +238,8 @@ var PDFFindController = (function PDFFindControllerClosure() {
}
// Update the matches count
if (matches.length > 0) {
this.matchCount += matches.length;
if (this.pageMatches[pageIndex].length > 0) {
this.matchCount += this.pageMatches[pageIndex].length;
this.updateUIResultsCount();
}
},
@ -233,6 +334,7 @@ var PDFFindController = (function PDFFindControllerClosure() {
this.resumePageIdx = null;
this.pageMatches = [];
this.matchCount = 0;
this.pageMatchesLength = null;
var self = this;
for (var i = 0; i < numPages; i++) {