@mlightcad/cad-pdf-plugin 1.8.0-preview.3 → 1.8.0-preview.4

package/lib/recognition/pdfTextMerge.js

@@ -0,0 +1,382 @@
+/**
+ * Merges recovered text lines (often single characters the clustering left
+ * apart) into one text block: groups them into paragraphs by baseline,
+ * orders them for reading and joins their texts.
+ *
+ * Pure geometry, like {@link clusterPdfTextStrokes}: unit-testable without
+ * the data model.
+ */
+var __rest = (this && this.__rest) || function (s, e) {
+    var t = {};
+    for (var p in s) if (Object.prototype.hasOwnProperty.call(s, p) && e.indexOf(p) < 0)
+        t[p] = s[p];
+    if (s != null && typeof Object.getOwnPropertySymbols === "function")
+        for (var i = 0, p = Object.getOwnPropertySymbols(s); i < p.length; i++) {
+            if (e.indexOf(p[i]) < 0 && Object.prototype.propertyIsEnumerable.call(s, p[i]))
+                t[p[i]] = s[p[i]];
+        }
+    return t;
+};
+var __read = (this && this.__read) || function (o, n) {
+    var m = typeof Symbol === "function" && o[Symbol.iterator];
+    if (!m) return o;
+    var i = m.call(o), r, ar = [], e;
+    try {
+        while ((n === void 0 || n-- > 0) && !(r = i.next()).done) ar.push(r.value);
+    }
+    catch (error) { e = { error: error }; }
+    finally {
+        try {
+            if (r && !r.done && (m = i["return"])) m.call(i);
+        }
+        finally { if (e) throw e.error; }
+    }
+    return ar;
+};
+var __spreadArray = (this && this.__spreadArray) || function (to, from, pack) {
+    if (pack || arguments.length === 2) for (var i = 0, l = from.length, ar; i < l; i++) {
+        if (ar || !(i in from)) {
+            if (!ar) ar = Array.prototype.slice.call(from, 0, i);
+            ar[i] = from[i];
+        }
+    }
+    return to.concat(ar || Array.prototype.slice.call(from));
+};
+var __values = (this && this.__values) || function(o) {
+    var s = typeof Symbol === "function" && Symbol.iterator, m = s && o[s], i = 0;
+    if (m) return m.call(o);
+    if (o && typeof o.length === "number") return {
+        next: function () {
+            if (o && i >= o.length) o = void 0;
+            return { value: o && o[i++], done: !o };
+        }
+    };
+    throw new TypeError(s ? "Object is not iterable." : "Symbol.iterator is not defined.");
+};
+/** Gap between two parts of a line (× height) that reads as a space. */
+var SPACE_GAP = 0.5;
+/** Baseline difference (× height) that starts a new paragraph. */
+var PARAGRAPH_GAP = 0.6;
+/**
+ * Merges lines into one block.
+ *
+ * The block direction is that of the part with the most glyphs: a single
+ * glyph's own direction is often a poor guess (an `I` looks vertical).
+ */
+export function mergePdfTextLines(parts, options) {
+    var e_1, _a;
+    if (options === void 0) { options = {}; }
+    var angle = dominantRotation(parts.map(function (part) { return part.line; }));
+    var boxes = parts.map(function (part, index) { return partBox(part, index, angle); });
+    var height = median(boxes.map(function (box) { return box.top - box.baseline; })) || 1e-9;
+    // Paragraphs: parts sharing a baseline, from top to bottom.
+    var byBaseline = __spreadArray([], __read(boxes), false).sort(function (a, b) { return b.baseline - a.baseline; });
+    var groups = [];
+    try {
+        for (var byBaseline_1 = __values(byBaseline), byBaseline_1_1 = byBaseline_1.next(); !byBaseline_1_1.done; byBaseline_1_1 = byBaseline_1.next()) {
+            var box = byBaseline_1_1.value;
+            var group = groups[groups.length - 1];
+            if (group &&
+                Math.abs(median(group.map(function (item) { return item.baseline; })) - box.baseline) <=
+                    PARAGRAPH_GAP * height) {
+                group.push(box);
+            }
+            else {
+                groups.push([box]);
+            }
+        }
+    }
+    catch (e_1_1) { e_1 = { error: e_1_1 }; }
+    finally {
+        try {
+            if (byBaseline_1_1 && !byBaseline_1_1.done && (_a = byBaseline_1.return)) _a.call(byBaseline_1);
+        }
+        finally { if (e_1) throw e_1.error; }
+    }
+    var paragraphs = groups.map(function (group) {
+        var e_2, _a;
+        var ordered = __spreadArray([], __read(group), false).sort(function (a, b) { return a.minU - b.minU; });
+        var spaces = spaceCounter(ordered, height, options.measure);
+        var text = '';
+        var previous;
+        try {
+            for (var ordered_1 = __values(ordered), ordered_1_1 = ordered_1.next(); !ordered_1_1.done; ordered_1_1 = ordered_1.next()) {
+                var box = ordered_1_1.value;
+                var piece = box.part.text.trim();
+                if (!piece)
+                    continue;
+                if (previous && text)
+                    text += ' '.repeat(spaces(previous, box));
+                text += piece;
+                previous = box;
+            }
+        }
+        catch (e_2_1) { e_2 = { error: e_2_1 }; }
+        finally {
+            try {
+                if (ordered_1_1 && !ordered_1_1.done && (_a = ordered_1.return)) _a.call(ordered_1);
+            }
+            finally { if (e_2) throw e_2.error; }
+        }
+        return {
+            parts: ordered.map(function (box) { return box.index; }),
+            baseline: median(group.map(function (box) { return box.baseline; })),
+            bottom: Math.min.apply(Math, __spreadArray([], __read(group.map(function (box) { return box.bottom; })), false)),
+            minU: Math.min.apply(Math, __spreadArray([], __read(group.map(function (box) { return box.minU; })), false)),
+            maxU: Math.max.apply(Math, __spreadArray([], __read(group.map(function (box) { return box.maxU; })), false)),
+            text: text,
+            boxes: ordered
+        };
+    });
+    var minU = Math.min.apply(Math, __spreadArray([], __read(paragraphs.map(function (item) { return item.minU; })), false));
+    var maxU = Math.max.apply(Math, __spreadArray([], __read(paragraphs.map(function (item) { return item.maxU; })), false));
+    var spacings = paragraphs
+        .slice(1)
+        .map(function (item, i) { return paragraphs[i].baseline - item.baseline; });
+    return {
+        line: {
+            strokes: parts.flatMap(function (part) { return part.line.strokes; }),
+            glyphs: paragraphs.flatMap(function (item) { return item.boxes.flatMap(function (box) { return box.glyphs; }); }),
+            rotation: angle,
+            position: toWorld(minU, paragraphs[0].baseline, angle),
+            height: height,
+            width: maxU - minU
+        },
+        paragraphs: paragraphs.map(function (_a) {
+            var _boxes = _a.boxes, rest = __rest(_a, ["boxes"]);
+            return rest;
+        }),
+        text: paragraphs.map(function (item) { return item.text; }).join('\n'),
+        lineSpacing: spacings.length > 0
+            ? spacings.reduce(function (sum, value) { return sum + value; }, 0) / spacings.length
+            : 0
+    };
+}
+/**
+ * Returns how many spaces go between two parts of a paragraph.
+ *
+ * With a measurer, the spaces fill the distance from the start of one part
+ * to the start of the next, less the advance of the first part's text, at
+ * the width factor that makes the parts' text span their ink (the median
+ * over the parts). Starts are used rather than the gap between the parts'
+ * boxes: a stroke the clustering put into the wrong part can widen a box,
+ * but hardly moves where a part begins.
+ *
+ * Without a measurer, a gap wider than half the text height is one space
+ * (also between CJK characters: letter-spaced titles such as "标   高"
+ * would otherwise lose their gaps).
+ */
+function spaceCounter(boxes, height, measure) {
+    var fallback = function (previous, next) {
+        return next.minU - previous.maxU > SPACE_GAP * height ? 1 : 0;
+    };
+    if (!measure)
+        return fallback;
+    var advance = function (text) {
+        var value = measure(text, height);
+        return value !== undefined && Number.isFinite(value) && value > 0
+            ? value
+            : undefined;
+    };
+    var fits = boxes.flatMap(function (box) {
+        var text = box.part.text.trim();
+        var natural = text ? advance(text) : undefined;
+        return natural ? [(box.maxU - box.minU) / natural] : [];
+    });
+    var widthFactor = median(fits);
+    var space = advance(' ');
+    if (!(widthFactor > 0) || !space)
+        return fallback;
+    return function (previous, next) {
+        var natural = advance(previous.part.text.trim());
+        if (natural === undefined)
+            return fallback(previous, next);
+        var free = next.minU - previous.minU - natural * widthFactor;
+        return Math.max(0, Math.round(free / (space * widthFactor)));
+    };
+}
+/**
+ * Whether two lines read as one: same direction (or a single glyph), similar
+ * height, the same baseline and close together along it.
+ *
+ * @param maxGap - Largest gap along the baseline, × text height.
+ */
+export function arePdfTextLinesAdjacent(a, b, maxGap) {
+    if (maxGap === void 0) { maxGap = 1.5; }
+    var angle = dominantRotation([a, b]);
+    if (!sameDirection(a, angle) || !sameDirection(b, angle))
+        return false;
+    var boxA = partBox({ line: a, text: '' }, 0, angle);
+    var boxB = partBox({ line: b, text: '' }, 1, angle);
+    var ha = boxA.top - boxA.baseline;
+    var hb = boxB.top - boxB.baseline;
+    var height = Math.max(ha, hb);
+    if (height <= 0)
+        return false;
+    // Single glyphs vary a lot in height (`-` vs `M`): only compare heights
+    // of real lines.
+    if (a.glyphs.length > 1 &&
+        b.glyphs.length > 1 &&
+        Math.min(ha, hb) < height * 0.6) {
+        return false;
+    }
+    if (Math.abs(boxA.baseline - boxB.baseline) > 0.3 * height)
+        return false;
+    var gap = Math.max(boxA.minU, boxB.minU) - Math.min(boxA.maxU, boxB.maxU);
+    return gap <= maxGap * height;
+}
+/** A line keeps the direction unless it is a single glyph (unreliable). */
+function sameDirection(line, angle) {
+    if (line.glyphs.length <= 1)
+        return true;
+    var diff = Math.abs(line.rotation - angle) % Math.PI;
+    if (diff > Math.PI / 2)
+        diff = Math.PI - diff;
+    return diff < (5 * Math.PI) / 180;
+}
+/**
+ * Direction of a merged block: that of the line with the most glyphs when
+ * one has several. Single glyphs only: their own directions are guesses, so
+ * the candidate (0, or one of theirs) giving the flattest block wins, like
+ * the minimum-height box of the clustering.
+ */
+function dominantRotation(lines) {
+    var e_3, _a, e_4, _b, e_5, _c;
+    var best = lines[0];
+    try {
+        for (var lines_1 = __values(lines), lines_1_1 = lines_1.next(); !lines_1_1.done; lines_1_1 = lines_1.next()) {
+            var line = lines_1_1.value;
+            if (line.glyphs.length > best.glyphs.length)
+                best = line;
+        }
+    }
+    catch (e_3_1) { e_3 = { error: e_3_1 }; }
+    finally {
+        try {
+            if (lines_1_1 && !lines_1_1.done && (_a = lines_1.return)) _a.call(lines_1);
+        }
+        finally { if (e_3) throw e_3.error; }
+    }
+    if (!best)
+        return 0;
+    if (best.glyphs.length > 1)
+        return best.rotation;
+    var points = lines.flatMap(function (line) {
+        return line.strokes.flatMap(function (stroke) { return stroke.points; });
+    });
+    var candidates = __spreadArray([0], __read(lines.map(function (line) { return line.rotation; })), false);
+    var bestAngle = 0;
+    var bestHeight = Infinity;
+    try {
+        for (var candidates_1 = __values(candidates), candidates_1_1 = candidates_1.next(); !candidates_1_1.done; candidates_1_1 = candidates_1.next()) {
+            var angle = candidates_1_1.value;
+            var minV = Infinity;
+            var maxV = -Infinity;
+            try {
+                for (var points_1 = (e_5 = void 0, __values(points)), points_1_1 = points_1.next(); !points_1_1.done; points_1_1 = points_1.next()) {
+                    var p = points_1_1.value;
+                    var v = toLocal(p, angle).v;
+                    minV = Math.min(minV, v);
+                    maxV = Math.max(maxV, v);
+                }
+            }
+            catch (e_5_1) { e_5 = { error: e_5_1 }; }
+            finally {
+                try {
+                    if (points_1_1 && !points_1_1.done && (_c = points_1.return)) _c.call(points_1);
+                }
+                finally { if (e_5) throw e_5.error; }
+            }
+            // Prefer the earlier candidate (0 first) unless clearly flatter.
+            if (maxV - minV < bestHeight - 1e-9) {
+                bestHeight = maxV - minV;
+                bestAngle = angle;
+            }
+        }
+    }
+    catch (e_4_1) { e_4 = { error: e_4_1 }; }
+    finally {
+        try {
+            if (candidates_1_1 && !candidates_1_1.done && (_b = candidates_1.return)) _b.call(candidates_1);
+        }
+        finally { if (e_4) throw e_4.error; }
+    }
+    return bestAngle;
+}
+/** Glyph boxes, extent and baseline of a part in the block frame. */
+function partBox(part, index, angle) {
+    var source = part.line.glyphs.length > 0
+        ? part.line.glyphs
+        : [{ strokes: part.line.strokes, minU: 0, maxU: 0, minV: 0, maxV: 0 }];
+    var glyphs = source.map(function (glyph) {
+        var e_6, _a, e_7, _b;
+        var minU = Infinity;
+        var maxU = -Infinity;
+        var minV = Infinity;
+        var maxV = -Infinity;
+        try {
+            for (var _c = __values(glyph.strokes), _d = _c.next(); !_d.done; _d = _c.next()) {
+                var stroke = _d.value;
+                try {
+                    for (var _e = (e_7 = void 0, __values(stroke.points)), _f = _e.next(); !_f.done; _f = _e.next()) {
+                        var p = _f.value;
+                        var _g = toLocal(p, angle), u = _g.u, v = _g.v;
+                        minU = Math.min(minU, u);
+                        maxU = Math.max(maxU, u);
+                        minV = Math.min(minV, v);
+                        maxV = Math.max(maxV, v);
+                    }
+                }
+                catch (e_7_1) { e_7 = { error: e_7_1 }; }
+                finally {
+                    try {
+                        if (_f && !_f.done && (_b = _e.return)) _b.call(_e);
+                    }
+                    finally { if (e_7) throw e_7.error; }
+                }
+            }
+        }
+        catch (e_6_1) { e_6 = { error: e_6_1 }; }
+        finally {
+            try {
+                if (_d && !_d.done && (_a = _c.return)) _a.call(_c);
+            }
+            finally { if (e_6) throw e_6.error; }
+        }
+        return { strokes: glyph.strokes, minU: minU, maxU: maxU, minV: minV, maxV: maxV };
+    });
+    var ordered = glyphs
+        .filter(function (glyph) { return Number.isFinite(glyph.minU); })
+        .sort(function (a, b) { return a.minU - b.minU; });
+    // Like the clustering: descenders hang below, most glyphs sit on the line.
+    var baseline = median(ordered.map(function (glyph) { return glyph.minV; }));
+    var tops = ordered.map(function (glyph) { return glyph.maxV; }).sort(function (a, b) { return a - b; });
+    return {
+        index: index,
+        part: part,
+        glyphs: ordered,
+        minU: Math.min.apply(Math, __spreadArray([], __read(ordered.map(function (glyph) { return glyph.minU; })), false)),
+        maxU: Math.max.apply(Math, __spreadArray([], __read(ordered.map(function (glyph) { return glyph.maxU; })), false)),
+        baseline: baseline,
+        bottom: Math.min.apply(Math, __spreadArray([], __read(ordered.map(function (glyph) { return glyph.minV; })), false)),
+        top: tops[Math.min(tops.length - 1, Math.floor(tops.length * 0.9))]
+    };
+}
+function median(values) {
+    if (values.length === 0)
+        return 0;
+    var sorted = __spreadArray([], __read(values), false).sort(function (a, b) { return a - b; });
+    var mid = Math.floor(sorted.length / 2);
+    return sorted.length % 2 ? sorted[mid] : (sorted[mid - 1] + sorted[mid]) / 2;
+}
+function toLocal(p, angle) {
+    var c = Math.cos(angle);
+    var s = Math.sin(angle);
+    return { u: p.x * c + p.y * s, v: -p.x * s + p.y * c };
+}
+function toWorld(u, v, angle) {
+    var c = Math.cos(angle);
+    var s = Math.sin(angle);
+    return { x: u * c - v * s, y: u * s + v * c };
+}
+//# sourceMappingURL=pdfTextMerge.js.map

package/lib/recognition/pdfTextMerge.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"pdfTextMerge.js","sourceRoot":"","sources":["../../src/recognition/pdfTextMerge.ts"],"names":[],"mappings":"AAAA;;;;;;;GAOG;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AA4DH,wEAAwE;AACxE,IAAM,SAAS,GAAG,GAAG,CAAA;AAErB,kEAAkE;AAClE,IAAM,aAAa,GAAG,GAAG,CAAA;AAazB;;;;;GAKG;AACH,MAAM,UAAU,iBAAiB,CAC/B,KAA6B,EAC7B,OAAiC;;IAAjC,wBAAA,EAAA,YAAiC;IAEjC,IAAM,KAAK,GAAG,gBAAgB,CAAC,KAAK,CAAC,GAAG,CAAC,UAAA,IAAI,IAAI,OAAA,IAAI,CAAC,IAAI,EAAT,CAAS,CAAC,CAAC,CAAA;IAC5D,IAAM,KAAK,GAAG,KAAK,CAAC,GAAG,CAAC,UAAC,IAAI,EAAE,KAAK,IAAK,OAAA,OAAO,CAAC,IAAI,EAAE,KAAK,EAAE,KAAK,CAAC,EAA3B,CAA2B,CAAC,CAAA;IACrE,IAAM,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,UAAA,GAAG,IAAI,OAAA,GAAG,CAAC,GAAG,GAAG,GAAG,CAAC,QAAQ,EAAtB,CAAsB,CAAC,CAAC,IAAI,IAAI,CAAA;IAEvE,4DAA4D;IAC5D,IAAM,UAAU,GAAG,yBAAI,KAAK,UAAE,IAAI,CAAC,UAAC,CAAC,EAAE,CAAC,IAAK,OAAA,CAAC,CAAC,QAAQ,GAAG,CAAC,CAAC,QAAQ,EAAvB,CAAuB,CAAC,CAAA;IACrE,IAAM,MAAM,GAAoB,EAAE,CAAA;;QAClC,KAAkB,IAAA,eAAA,SAAA,UAAU,CAAA,sCAAA,8DAAE,CAAC;YAA1B,IAAM,GAAG,uBAAA;YACZ,IAAM,KAAK,GAAG,MAAM,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,CAAA;YACvC,IACE,KAAK;gBACL,IAAI,CAAC,GAAG,CAAC,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,UAAA,IAAI,IAAI,OAAA,IAAI,CAAC,QAAQ,EAAb,CAAa,CAAC,CAAC,GAAG,GAAG,CAAC,QAAQ,CAAC;oBAC/D,aAAa,GAAG,MAAM,EACxB,CAAC;gBACD,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAA;YACjB,CAAC;iBAAM,CAAC;gBACN,MAAM,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAA;YACpB,CAAC;QACH,CAAC;;;;;;;;;IACD,IAAM,UAAU,GAAG,MAAM,CAAC,GAAG,CAAC,UAAA,KAAK;;QACjC,IAAM,OAAO,GAAG,yBAAI,KAAK,UAAE,IAAI,CAAC,UAAC,CAAC,EAAE,CAAC,IAAK,OAAA,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,EAAf,CAAe,CAAC,CAAA;QAC1D,IAAM,MAAM,GAAG,YAAY,CAAC,OAAO,EAAE,MAAM,EAAE,OAAO,CAAC,OAAO,CAAC,CAAA;QAC7D,IAAI,IAAI,GAAG,EAAE,CAAA;QACb,IAAI,QAAiC,CAAA;;YACrC,KAAkB,IAAA,YAAA,SAAA,OAAO,CAAA,gCAAA,qDAAE,CAAC;gBAAvB,IAAM,GAAG,oBAAA;gBACZ,IAAM,KAAK,GAAG,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,CAAA;gBAClC,IAAI,CAAC,KAAK;oBAAE,SAAQ;gBACpB,IAAI,QAAQ,IAAI,IAAI;oBAAE,IAAI,IAAI,GAAG,CAAC,MAAM,CAAC,MAAM,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC,CAAA;gBAC/D,IAAI,IAAI,KAAK,CAAA;gBACb,QAAQ,GAAG,GAAG,CAAA;YAChB,CAAC;;;;;;;;;QACD,OAAO;YACL,KAAK,EAAE,OAAO,CAAC,GAAG,CAAC,UAAA,GAAG,IAAI,OAAA,GAAG,CAAC,KAAK,EAAT,CAAS,CAAC;YACpC,QAAQ,EAAE,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,UAAA,GAAG,IAAI,OAAA,GAAG,CAAC,QAAQ,EAAZ,CAAY,CAAC,CAAC;YAChD,MAAM,EAAE,IAAI,CAAC,GAAG,OAAR,IAAI,2BAAQ,KAAK,CAAC,GAAG,CAAC,UAAA,GAAG,IAAI,OAAA,GAAG,CAAC,MAAM,EAAV,CAAU,CAAC,UAAC;YACjD,IAAI,EAAE,IAAI,CAAC,GAAG,OAAR,IAAI,2BAAQ,KAAK,CAAC,GAAG,CAAC,UAAA,GAAG,IAAI,OAAA,GAAG,CAAC,IAAI,EAAR,CAAQ,CAAC,UAAC;YAC7C,IAAI,EAAE,IAAI,CAAC,GAAG,OAAR,IAAI,2BAAQ,KAAK,CAAC,GAAG,CAAC,UAAA,GAAG,IAAI,OAAA,GAAG,CAAC,IAAI,EAAR,CAAQ,CAAC,UAAC;YAC7C,IAAI,MAAA;YACJ,KAAK,EAAE,OAAO;SACf,CAAA;IACH,CAAC,CAAC,CAAA;IAEF,IAAM,IAAI,GAAG,IAAI,CAAC,GAAG,OAAR,IAAI,2BAAQ,UAAU,CAAC,GAAG,CAAC,UAAA,IAAI,IAAI,OAAA,IAAI,CAAC,IAAI,EAAT,CAAS,CAAC,UAAC,CAAA;IAC3D,IAAM,IAAI,GAAG,IAAI,CAAC,GAAG,OAAR,IAAI,2BAAQ,UAAU,CAAC,GAAG,CAAC,UAAA,IAAI,IAAI,OAAA,IAAI,CAAC,IAAI,EAAT,CAAS,CAAC,UAAC,CAAA;IAC3D,IAAM,QAAQ,GAAG,UAAU;SACxB,KAAK,CAAC,CAAC,CAAC;SACR,GAAG,CAAC,UAAC,IAAI,EAAE,CAAC,IAAK,OAAA,UAAU,CAAC,CAAC,CAAC,CAAC,QAAQ,GAAG,IAAI,CAAC,QAAQ,EAAtC,CAAsC,CAAC,CAAA;IAC3D,OAAO;QACL,IAAI,EAAE;YACJ,OAAO,EAAE,KAAK,CAAC,OAAO,CAAC,UAAA,IAAI,IAAI,OAAA,IAAI,CAAC,IAAI,CAAC,OAAO,EAAjB,CAAiB,CAAC;YACjD,MAAM,EAAE,UAAU,CAAC,OAAO,CAAC,UAAA,IAAI,IAAI,OAAA,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,UAAA,GAAG,IAAI,OAAA,GAAG,CAAC,MAAM,EAAV,CAAU,CAAC,EAArC,CAAqC,CAAC;YACzE,QAAQ,EAAE,KAAK;YACf,QAAQ,EAAE,OAAO,CAAC,IAAI,EAAE,UAAU,CAAC,CAAC,CAAC,CAAC,QAAQ,EAAE,KAAK,CAAC;YACtD,MAAM,QAAA;YACN,KAAK,EAAE,IAAI,GAAG,IAAI;SACnB;QACD,UAAU,EAAE,UAAU,CAAC,GAAG,CAAC,UAAC,EAA0B;YAAxB,IAAO,MAAM,WAAA,EAAK,IAAI,cAAxB,SAA0B,CAAF;YAAO,OAAA,IAAI,CAAA;SAAA,CAAC;QAChE,IAAI,EAAE,UAAU,CAAC,GAAG,CAAC,UAAA,IAAI,IAAI,OAAA,IAAI,CAAC,IAAI,EAAT,CAAS,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC;QAClD,WAAW,EACT,QAAQ,CAAC,MAAM,GAAG,CAAC;YACjB,CAAC,CAAC,QAAQ,CAAC,MAAM,CAAC,UAAC,GAAG,EAAE,KAAK,IAAK,OAAA,GAAG,GAAG,KAAK,EAAX,CAAW,EAAE,CAAC,CAAC,GAAG,QAAQ,CAAC,MAAM;YACnE,CAAC,CAAC,CAAC;KACR,CAAA;AACH,CAAC;AAED;;;;;;;;;;;;;GAaG;AACH,SAAS,YAAY,CACnB,KAAoB,EACpB,MAAc,EACd,OAAwC;IAExC,IAAM,QAAQ,GAAG,UAAC,QAAqB,EAAE,IAAiB;QACxD,OAAA,IAAI,CAAC,IAAI,GAAG,QAAQ,CAAC,IAAI,GAAG,SAAS,GAAG,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAAtD,CAAsD,CAAA;IACxD,IAAI,CAAC,OAAO;QAAE,OAAO,QAAQ,CAAA;IAC7B,IAAM,OAAO,GAAG,UAAC,IAAY;QAC3B,IAAM,KAAK,GAAG,OAAO,CAAC,IAAI,EAAE,MAAM,CAAC,CAAA;QACnC,OAAO,KAAK,KAAK,SAAS,IAAI,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,IAAI,KAAK,GAAG,CAAC;YAC/D,CAAC,CAAC,KAAK;YACP,CAAC,CAAC,SAAS,CAAA;IACf,CAAC,CAAA;IACD,IAAM,IAAI,GAAG,KAAK,CAAC,OAAO,CAAC,UAAA,GAAG;QAC5B,IAAM,IAAI,GAAG,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,CAAA;QACjC,IAAM,OAAO,GAAG,IAAI,CAAC,CAAC,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,SAAS,CAAA;QAChD,OAAO,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,IAAI,GAAG,GAAG,CAAC,IAAI,CAAC,GAAG,OAAO,CAAC,CAAC,CAAC,CAAC,EAAE,CAAA;IACzD,CAAC,CAAC,CAAA;IACF,IAAM,WAAW,GAAG,MAAM,CAAC,IAAI,CAAC,CAAA;IAChC,IAAM,KAAK,GAAG,OAAO,CAAC,GAAG,CAAC,CAAA;IAC1B,IAAI,CAAC,CAAC,WAAW,GAAG,CAAC,CAAC,IAAI,CAAC,KAAK;QAAE,OAAO,QAAQ,CAAA;IACjD,OAAO,UAAC,QAAQ,EAAE,IAAI;QACpB,IAAM,OAAO,GAAG,OAAO,CAAC,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,CAAA;QAClD,IAAI,OAAO,KAAK,SAAS;YAAE,OAAO,QAAQ,CAAC,QAAQ,EAAE,IAAI,CAAC,CAAA;QAC1D,IAAM,IAAI,GAAG,IAAI,CAAC,IAAI,GAAG,QAAQ,CAAC,IAAI,GAAG,OAAO,GAAG,WAAW,CAAA;QAC9D,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,IAAI,GAAG,CAAC,KAAK,GAAG,WAAW,CAAC,CAAC,CAAC,CAAA;IAC9D,CAAC,CAAA;AACH,CAAC;AAED;;;;;GAKG;AACH,MAAM,UAAU,uBAAuB,CACrC,CAAuB,EACvB,CAAuB,EACvB,MAAY;IAAZ,uBAAA,EAAA,YAAY;IAEZ,IAAM,KAAK,GAAG,gBAAgB,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAA;IACtC,IAAI,CAAC,aAAa,CAAC,CAAC,EAAE,KAAK,CAAC,IAAI,CAAC,aAAa,CAAC,CAAC,EAAE,KAAK,CAAC;QAAE,OAAO,KAAK,CAAA;IACtE,IAAM,IAAI,GAAG,OAAO,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,EAAE,EAAE,EAAE,CAAC,EAAE,KAAK,CAAC,CAAA;IACrD,IAAM,IAAI,GAAG,OAAO,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,EAAE,EAAE,EAAE,CAAC,EAAE,KAAK,CAAC,CAAA;IACrD,IAAM,EAAE,GAAG,IAAI,CAAC,GAAG,GAAG,IAAI,CAAC,QAAQ,CAAA;IACnC,IAAM,EAAE,GAAG,IAAI,CAAC,GAAG,GAAG,IAAI,CAAC,QAAQ,CAAA;IACnC,IAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,EAAE,EAAE,EAAE,CAAC,CAAA;IAC/B,IAAI,MAAM,IAAI,CAAC;QAAE,OAAO,KAAK,CAAA;IAC7B,wEAAwE;IACxE,iBAAiB;IACjB,IACE,CAAC,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC;QACnB,CAAC,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC;QACnB,IAAI,CAAC,GAAG,CAAC,EAAE,EAAE,EAAE,CAAC,GAAG,MAAM,GAAG,GAAG,EAC/B,CAAC;QACD,OAAO,KAAK,CAAA;IACd,CAAC;IACD,IAAI,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,QAAQ,GAAG,IAAI,CAAC,QAAQ,CAAC,GAAG,GAAG,GAAG,MAAM;QAAE,OAAO,KAAK,CAAA;IACxE,IAAM,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,EAAE,IAAI,CAAC,IAAI,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,EAAE,IAAI,CAAC,IAAI,CAAC,CAAA;IAC3E,OAAO,GAAG,IAAI,MAAM,GAAG,MAAM,CAAA;AAC/B,CAAC;AAED,2EAA2E;AAC3E,SAAS,aAAa,CAAC,IAA0B,EAAE,KAAa;IAC9D,IAAI,IAAI,CAAC,MAAM,CAAC,MAAM,IAAI,CAAC;QAAE,OAAO,IAAI,CAAA;IACxC,IAAI,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,QAAQ,GAAG,KAAK,CAAC,GAAG,IAAI,CAAC,EAAE,CAAA;IACpD,IAAI,IAAI,GAAG,IAAI,CAAC,EAAE,GAAG,CAAC;QAAE,IAAI,GAAG,IAAI,CAAC,EAAE,GAAG,IAAI,CAAA;IAC7C,OAAO,IAAI,GAAG,CAAC,CAAC,GAAG,IAAI,CAAC,EAAE,CAAC,GAAG,GAAG,CAAA;AACnC,CAAC;AAED;;;;;GAKG;AACH,SAAS,gBAAgB,CAAC,KAA6B;;IACrD,IAAI,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,CAAA;;QACnB,KAAmB,IAAA,UAAA,SAAA,KAAK,CAAA,4BAAA,+CAAE,CAAC;YAAtB,IAAM,IAAI,kBAAA;YACb,IAAI,IAAI,CAAC,MAAM,CAAC,MAAM,GAAG,IAAI,CAAC,MAAM,CAAC,MAAM;gBAAE,IAAI,GAAG,IAAI,CAAA;QAC1D,CAAC;;;;;;;;;IACD,IAAI,CAAC,IAAI;QAAE,OAAO,CAAC,CAAA;IACnB,IAAI,IAAI,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC;QAAE,OAAO,IAAI,CAAC,QAAQ,CAAA;IAChD,IAAM,MAAM,GAAG,KAAK,CAAC,OAAO,CAAC,UAAA,IAAI;QAC/B,OAAA,IAAI,CAAC,OAAO,CAAC,OAAO,CAAC,UAAA,MAAM,IAAI,OAAA,MAAM,CAAC,MAAM,EAAb,CAAa,CAAC;IAA7C,CAA6C,CAC9C,CAAA;IACD,IAAM,UAAU,kBAAI,CAAC,UAAK,KAAK,CAAC,GAAG,CAAC,UAAA,IAAI,IAAI,OAAA,IAAI,CAAC,QAAQ,EAAb,CAAa,CAAC,SAAC,CAAA;IAC3D,IAAI,SAAS,GAAG,CAAC,CAAA;IACjB,IAAI,UAAU,GAAG,QAAQ,CAAA;;QACzB,KAAoB,IAAA,eAAA,SAAA,UAAU,CAAA,sCAAA,8DAAE,CAAC;YAA5B,IAAM,KAAK,uBAAA;YACd,IAAI,IAAI,GAAG,QAAQ,CAAA;YACnB,IAAI,IAAI,GAAG,CAAC,QAAQ,CAAA;;gBACpB,KAAgB,IAAA,0BAAA,SAAA,MAAM,CAAA,CAAA,8BAAA,kDAAE,CAAC;oBAApB,IAAM,CAAC,mBAAA;oBACF,IAAA,CAAC,GAAK,OAAO,CAAC,CAAC,EAAE,KAAK,CAAC,EAAtB,CAAsB;oBAC/B,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC,CAAC,CAAA;oBACxB,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC,CAAC,CAAA;gBAC1B,CAAC;;;;;;;;;YACD,iEAAiE;YACjE,IAAI,IAAI,GAAG,IAAI,GAAG,UAAU,GAAG,IAAI,EAAE,CAAC;gBACpC,UAAU,GAAG,IAAI,GAAG,IAAI,CAAA;gBACxB,SAAS,GAAG,KAAK,CAAA;YACnB,CAAC;QACH,CAAC;;;;;;;;;IACD,OAAO,SAAS,CAAA;AAClB,CAAC;AAED,qEAAqE;AACrE,SAAS,OAAO,CACd,IAA0B,EAC1B,KAAa,EACb,KAAa;IAEb,IAAM,MAAM,GACV,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC;QACzB,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,MAAM;QAClB,CAAC,CAAC,CAAC,EAAE,OAAO,EAAE,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,CAAC,CAAA;IAC1E,IAAM,MAAM,GAAG,MAAM,CAAC,GAAG,CAAC,UAAA,KAAK;;QAC7B,IAAI,IAAI,GAAG,QAAQ,CAAA;QACnB,IAAI,IAAI,GAAG,CAAC,QAAQ,CAAA;QACpB,IAAI,IAAI,GAAG,QAAQ,CAAA;QACnB,IAAI,IAAI,GAAG,CAAC,QAAQ,CAAA;;YACpB,KAAqB,IAAA,KAAA,SAAA,KAAK,CAAC,OAAO,CAAA,gBAAA,4BAAE,CAAC;gBAAhC,IAAM,MAAM,WAAA;;oBACf,KAAgB,IAAA,oBAAA,SAAA,MAAM,CAAC,MAAM,CAAA,CAAA,gBAAA,4BAAE,CAAC;wBAA3B,IAAM,CAAC,WAAA;wBACJ,IAAA,KAAW,OAAO,CAAC,CAAC,EAAE,KAAK,CAAC,EAA1B,CAAC,OAAA,EAAE,CAAC,OAAsB,CAAA;wBAClC,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC,CAAC,CAAA;wBACxB,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC,CAAC,CAAA;wBACxB,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC,CAAC,CAAA;wBACxB,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC,CAAC,CAAA;oBAC1B,CAAC;;;;;;;;;YACH,CAAC;;;;;;;;;QACD,OAAO,EAAE,OAAO,EAAE,KAAK,CAAC,OAAO,EAAE,IAAI,MAAA,EAAE,IAAI,MAAA,EAAE,IAAI,MAAA,EAAE,IAAI,MAAA,EAAE,CAAA;IAC3D,CAAC,CAAC,CAAA;IACF,IAAM,OAAO,GAAG,MAAM;SACnB,MAAM,CAAC,UAAA,KAAK,IAAI,OAAA,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,IAAI,CAAC,EAA3B,CAA2B,CAAC;SAC5C,IAAI,CAAC,UAAC,CAAC,EAAE,CAAC,IAAK,OAAA,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,EAAf,CAAe,CAAC,CAAA;IAClC,2EAA2E;IAC3E,IAAM,QAAQ,GAAG,MAAM,CAAC,OAAO,CAAC,GAAG,CAAC,UAAA,KAAK,IAAI,OAAA,KAAK,CAAC,IAAI,EAAV,CAAU,CAAC,CAAC,CAAA;IACzD,IAAM,IAAI,GAAG,OAAO,CAAC,GAAG,CAAC,UAAA,KAAK,IAAI,OAAA,KAAK,CAAC,IAAI,EAAV,CAAU,CAAC,CAAC,IAAI,CAAC,UAAC,CAAC,EAAE,CAAC,IAAK,OAAA,CAAC,GAAG,CAAC,EAAL,CAAK,CAAC,CAAA;IACnE,OAAO;QACL,KAAK,OAAA;QACL,IAAI,MAAA;QACJ,MAAM,EAAE,OAAO;QACf,IAAI,EAAE,IAAI,CAAC,GAAG,OAAR,IAAI,2BAAQ,OAAO,CAAC,GAAG,CAAC,UAAA,KAAK,IAAI,OAAA,KAAK,CAAC,IAAI,EAAV,CAAU,CAAC,UAAC;QACnD,IAAI,EAAE,IAAI,CAAC,GAAG,OAAR,IAAI,2BAAQ,OAAO,CAAC,GAAG,CAAC,UAAA,KAAK,IAAI,OAAA,KAAK,CAAC,IAAI,EAAV,CAAU,CAAC,UAAC;QACnD,QAAQ,UAAA;QACR,MAAM,EAAE,IAAI,CAAC,GAAG,OAAR,IAAI,2BAAQ,OAAO,CAAC,GAAG,CAAC,UAAA,KAAK,IAAI,OAAA,KAAK,CAAC,IAAI,EAAV,CAAU,CAAC,UAAC;QACrD,GAAG,EAAE,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,MAAM,GAAG,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,MAAM,GAAG,GAAG,CAAC,CAAC,CAAC;KACpE,CAAA;AACH,CAAC;AAED,SAAS,MAAM,CAAC,MAAgB;IAC9B,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,CAAC,CAAA;IACjC,IAAM,MAAM,GAAG,yBAAI,MAAM,UAAE,IAAI,CAAC,UAAC,CAAC,EAAE,CAAC,IAAK,OAAA,CAAC,GAAG,CAAC,EAAL,CAAK,CAAC,CAAA;IAChD,IAAM,GAAG,GAAG,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,CAAA;IACzC,OAAO,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,GAAG,CAAC,CAAC,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAA;AAC9E,CAAC;AAED,SAAS,OAAO,CAAC,CAAc,EAAE,KAAa;IAC5C,IAAM,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,CAAA;IACzB,IAAM,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,CAAA;IACzB,OAAO,EAAE,CAAC,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,EAAE,CAAA;AACxD,CAAC;AAED,SAAS,OAAO,CAAC,CAAS,EAAE,CAAS,EAAE,KAAa;IAClD,IAAM,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,CAAA;IACzB,IAAM,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,CAAA;IACzB,OAAO,EAAE,CAAC,EAAE,CAAC,GAAG,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,CAAC,GAAG,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,CAAA;AAC/C,CAAC"}

package/lib/recognition/pdfTextRunMerge.d.ts

@@ -0,0 +1,28 @@
+/**
+ * Joins consecutive PDF text runs that continue the same line.
+ *
+ * PDF producers often emit one line as several `showText` runs (kerning,
+ * font switches, per-word positioning). Importing each run as its own
+ * `AcDbText` leaves a line broken into fragments that are hard to edit.
+ */
+import type { PdfImportedText } from '../pdfVectorImport';
+/** Tuning for {@link mergePdfTextRuns}. */
+export interface PdfTextRunMergeOptions {
+    /** Largest gap between runs, as a fraction of text height. Default 0.6. */
+    maxGap?: number;
+    /** Gap that becomes a space, as a fraction of text height. Default 0.2. */
+    spaceGap?: number;
+}
+/**
+ * Merges each run into the previous one when both sit on the same baseline
+ * with the same style and the gap between them is small. Only consecutive
+ * runs are joined, so content-stream order (normally reading order) decides.
+ *
+ * Runs without a known advance width are never merged into.
+ *
+ * @param runs - Text runs in content-stream order.
+ * @param options - Gap thresholds.
+ * @returns New runs; inputs are not modified.
+ */
+export declare function mergePdfTextRuns(runs: readonly PdfImportedText[], options?: PdfTextRunMergeOptions): PdfImportedText[];
+//# sourceMappingURL=pdfTextRunMerge.d.ts.map

package/lib/recognition/pdfTextRunMerge.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"pdfTextRunMerge.d.ts","sourceRoot":"","sources":["../../src/recognition/pdfTextRunMerge.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AAEH,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,oBAAoB,CAAA;AAEzD,2CAA2C;AAC3C,MAAM,WAAW,sBAAsB;IACrC,2EAA2E;IAC3E,MAAM,CAAC,EAAE,MAAM,CAAA;IACf,2EAA2E;IAC3E,QAAQ,CAAC,EAAE,MAAM,CAAA;CAClB;AAED;;;;;;;;;;GAUG;AACH,wBAAgB,gBAAgB,CAC9B,IAAI,EAAE,SAAS,eAAe,EAAE,EAChC,OAAO,GAAE,sBAA2B,GACnC,eAAe,EAAE,CAgCnB"}

package/lib/recognition/pdfTextRunMerge.js

@@ -0,0 +1,90 @@
+/**
+ * Joins consecutive PDF text runs that continue the same line.
+ *
+ * PDF producers often emit one line as several `showText` runs (kerning,
+ * font switches, per-word positioning). Importing each run as its own
+ * `AcDbText` leaves a line broken into fragments that are hard to edit.
+ */
+var __assign = (this && this.__assign) || function () {
+    __assign = Object.assign || function(t) {
+        for (var s, i = 1, n = arguments.length; i < n; i++) {
+            s = arguments[i];
+            for (var p in s) if (Object.prototype.hasOwnProperty.call(s, p))
+                t[p] = s[p];
+        }
+        return t;
+    };
+    return __assign.apply(this, arguments);
+};
+var __values = (this && this.__values) || function(o) {
+    var s = typeof Symbol === "function" && Symbol.iterator, m = s && o[s], i = 0;
+    if (m) return m.call(o);
+    if (o && typeof o.length === "number") return {
+        next: function () {
+            if (o && i >= o.length) o = void 0;
+            return { value: o && o[i++], done: !o };
+        }
+    };
+    throw new TypeError(s ? "Object is not iterable." : "Symbol.iterator is not defined.");
+};
+/**
+ * Merges each run into the previous one when both sit on the same baseline
+ * with the same style and the gap between them is small. Only consecutive
+ * runs are joined, so content-stream order (normally reading order) decides.
+ *
+ * Runs without a known advance width are never merged into.
+ *
+ * @param runs - Text runs in content-stream order.
+ * @param options - Gap thresholds.
+ * @returns New runs; inputs are not modified.
+ */
+export function mergePdfTextRuns(runs, options) {
+    var e_1, _a;
+    var _b, _c;
+    if (options === void 0) { options = {}; }
+    var maxGap = (_b = options.maxGap) !== null && _b !== void 0 ? _b : 0.6;
+    var spaceGap = (_c = options.spaceGap) !== null && _c !== void 0 ? _c : 0.2;
+    var out = [];
+    try {
+        for (var runs_1 = __values(runs), runs_1_1 = runs_1.next(); !runs_1_1.done; runs_1_1 = runs_1.next()) {
+            var run = runs_1_1.value;
+            var prev = out[out.length - 1];
+            if (prev && prev.width !== undefined && canJoin(prev, run)) {
+                var dirX = Math.cos(prev.rotation);
+                var dirY = Math.sin(prev.rotation);
+                var dx = run.position.x - prev.position.x;
+                var dy = run.position.y - prev.position.y;
+                var along = dx * dirX + dy * dirY;
+                var across = Math.abs(-dx * dirY + dy * dirX);
+                var gap = along - prev.width;
+                var h = prev.height;
+                if (across <= h * 0.1 && gap >= -h * 0.2 && gap <= h * maxGap) {
+                    var needsSpace = gap > h * spaceGap && !/\s$/.test(prev.text) && !/^\s/.test(run.text);
+                    out[out.length - 1] = __assign(__assign({}, prev), { text: prev.text + (needsSpace ? ' ' : '') + run.text, width: run.width !== undefined
+                            ? Math.max(prev.width, along + run.width)
+                            : undefined });
+                    continue;
+                }
+            }
+            out.push(__assign({}, run));
+        }
+    }
+    catch (e_1_1) { e_1 = { error: e_1_1 }; }
+    finally {
+        try {
+            if (runs_1_1 && !runs_1_1.done && (_a = runs_1.return)) _a.call(runs_1);
+        }
+        finally { if (e_1) throw e_1.error; }
+    }
+    return out;
+}
+function canJoin(a, b) {
+    var rotationDiff = Math.abs(Math.atan2(Math.sin(a.rotation - b.rotation), Math.cos(a.rotation - b.rotation)));
+    return (a.layerName === b.layerName &&
+        a.color === b.color &&
+        a.fontId === b.fontId &&
+        Math.abs(a.height - b.height) <= Math.max(a.height, b.height) * 0.01 &&
+        Math.abs(a.widthFactor - b.widthFactor) <= 0.01 &&
+        rotationDiff <= (0.5 * Math.PI) / 180);
+}
+//# sourceMappingURL=pdfTextRunMerge.js.map

package/lib/recognition/pdfTextRunMerge.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"pdfTextRunMerge.js","sourceRoot":"","sources":["../../src/recognition/pdfTextRunMerge.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;;;;;;;;;;;;;;;;;;;;;;;AAYH;;;;;;;;;;GAUG;AACH,MAAM,UAAU,gBAAgB,CAC9B,IAAgC,EAChC,OAAoC;;;IAApC,wBAAA,EAAA,YAAoC;IAEpC,IAAM,MAAM,GAAG,MAAA,OAAO,CAAC,MAAM,mCAAI,GAAG,CAAA;IACpC,IAAM,QAAQ,GAAG,MAAA,OAAO,CAAC,QAAQ,mCAAI,GAAG,CAAA;IACxC,IAAM,GAAG,GAAsB,EAAE,CAAA;;QACjC,KAAkB,IAAA,SAAA,SAAA,IAAI,CAAA,0BAAA,4CAAE,CAAC;YAApB,IAAM,GAAG,iBAAA;YACZ,IAAM,IAAI,GAAG,GAAG,CAAC,GAAG,CAAC,MAAM,GAAG,CAAC,CAAC,CAAA;YAChC,IAAI,IAAI,IAAI,IAAI,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,CAAC,IAAI,EAAE,GAAG,CAAC,EAAE,CAAC;gBAC3D,IAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAA;gBACpC,IAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAA;gBACpC,IAAM,EAAE,GAAG,GAAG,CAAC,QAAQ,CAAC,CAAC,GAAG,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAA;gBAC3C,IAAM,EAAE,GAAG,GAAG,CAAC,QAAQ,CAAC,CAAC,GAAG,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAA;gBAC3C,IAAM,KAAK,GAAG,EAAE,GAAG,IAAI,GAAG,EAAE,GAAG,IAAI,CAAA;gBACnC,IAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,GAAG,IAAI,GAAG,EAAE,GAAG,IAAI,CAAC,CAAA;gBAC/C,IAAM,GAAG,GAAG,KAAK,GAAG,IAAI,CAAC,KAAK,CAAA;gBAC9B,IAAM,CAAC,GAAG,IAAI,CAAC,MAAM,CAAA;gBACrB,IAAI,MAAM,IAAI,CAAC,GAAG,GAAG,IAAI,GAAG,IAAI,CAAC,CAAC,GAAG,GAAG,IAAI,GAAG,IAAI,CAAC,GAAG,MAAM,EAAE,CAAC;oBAC9D,IAAM,UAAU,GACd,GAAG,GAAG,CAAC,GAAG,QAAQ,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,CAAA;oBACvE,GAAG,CAAC,GAAG,CAAC,MAAM,GAAG,CAAC,CAAC,yBACd,IAAI,KACP,IAAI,EAAE,IAAI,CAAC,IAAI,GAAG,CAAC,UAAU,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,GAAG,GAAG,CAAC,IAAI,EACpD,KAAK,EACH,GAAG,CAAC,KAAK,KAAK,SAAS;4BACrB,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,KAAK,EAAE,KAAK,GAAG,GAAG,CAAC,KAAK,CAAC;4BACzC,CAAC,CAAC,SAAS,GAChB,CAAA;oBACD,SAAQ;gBACV,CAAC;YACH,CAAC;YACD,GAAG,CAAC,IAAI,cAAM,GAAG,EAAG,CAAA;QACtB,CAAC;;;;;;;;;IACD,OAAO,GAAG,CAAA;AACZ,CAAC;AAED,SAAS,OAAO,CAAC,CAAkB,EAAE,CAAkB;IACrD,IAAM,YAAY,GAAG,IAAI,CAAC,GAAG,CAC3B,IAAI,CAAC,KAAK,CACR,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,QAAQ,GAAG,CAAC,CAAC,QAAQ,CAAC,EACjC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,QAAQ,GAAG,CAAC,CAAC,QAAQ,CAAC,CAClC,CACF,CAAA;IACD,OAAO,CACL,CAAC,CAAC,SAAS,KAAK,CAAC,CAAC,SAAS;QAC3B,CAAC,CAAC,KAAK,KAAK,CAAC,CAAC,KAAK;QACnB,CAAC,CAAC,MAAM,KAAK,CAAC,CAAC,MAAM;QACrB,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,MAAM,CAAC,IAAI,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,MAAM,EAAE,CAAC,CAAC,MAAM,CAAC,GAAG,IAAI;QACpE,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,WAAW,GAAG,CAAC,CAAC,WAAW,CAAC,IAAI,IAAI;QAC/C,YAAY,IAAI,CAAC,GAAG,GAAG,IAAI,CAAC,EAAE,CAAC,GAAG,GAAG,CACtC,CAAA;AACH,CAAC"}

package/lib/register.d.ts

@@ -7,9 +7,15 @@ export declare const PDF_PLUGIN_NAME = "PdfPlugin";
  * - `-cpdf` — export drawing to PDF via command-line prompts
  * - `cpdf` — same as `-cpdf` when no UI dialog is registered; in cad-viewer,
  *   `cpdf` opens the export dialog and loads this plugin on confirm via `-cpdf`
- * - `ipdf` — import vector geometry from PDF
+ * - `pdfimport` — import vector geometry from PDF through the Import PDF dialog
+ * - `-pdfimport` — same import with command-line prompts
+ * - `pdfscale` — calibrate imported PDF geometry from a known distance
+ * - `pdftext` — convert exploded PDF text strokes back into text
+ * - `pdfblock` — turn repeated PDF geometry back into a block
+ * - `pdfdim` — turn exploded PDF dimensions back into dimensions
+ * - `pdfhatch` — turn the loose strokes of a PDF hatch back into a hatch
  */
-export declare const PDF_PLUGIN_TRIGGERS: readonly ["-cpdf", "cpdf", "ipdf"];
+export declare const PDF_PLUGIN_TRIGGERS: readonly ["-cpdf", "cpdf", "pdfimport", "-pdfimport", "pdfscale", "pdftext", "pdfblock", "pdfdim", "pdfhatch"];
 /**
  * Registers the PDF plugin for lazy loading.
  *

package/lib/register.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"register.d.ts","sourceRoot":"","sources":["../src/register.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,iBAAiB,EAAE,MAAM,8BAA8B,CAAA;AAErE,8CAA8C;AAC9C,eAAO,MAAM,eAAe,cAAc,CAAA;AAE1C;;;;;;;GAOG;AACH,eAAO,MAAM,mBAAmB,oCAAqC,CAAA;AAErE;;;;;;;GAOG;AACH,wBAAgB,qBAAqB,CACnC,aAAa,EAAE,iBAAiB,EAChC,OAAO,GAAE;IAAE,aAAa,CAAC,EAAE,OAAO,CAAA;CAAO,GACxC,IAAI,CAcN"}
+{"version":3,"file":"register.d.ts","sourceRoot":"","sources":["../src/register.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,iBAAiB,EAAE,MAAM,8BAA8B,CAAA;AAErE,8CAA8C;AAC9C,eAAO,MAAM,eAAe,cAAc,CAAA;AAE1C;;;;;;;;;;;;;GAaG;AACH,eAAO,MAAM,mBAAmB,gHAUtB,CAAA;AAaV;;;;;;;GAOG;AACH,wBAAgB,qBAAqB,CACnC,aAAa,EAAE,iBAAiB,EAChC,OAAO,GAAE;IAAE,aAAa,CAAC,EAAE,OAAO,CAAA;CAAO,GACxC,IAAI,CAcN"}

package/lib/register.js

@@ -67,9 +67,35 @@ export var PDF_PLUGIN_NAME = 'PdfPlugin';
  * - `-cpdf` — export drawing to PDF via command-line prompts
  * - `cpdf` — same as `-cpdf` when no UI dialog is registered; in cad-viewer,
  *   `cpdf` opens the export dialog and loads this plugin on confirm via `-cpdf`
- * - `ipdf` — import vector geometry from PDF
+ * - `pdfimport` — import vector geometry from PDF through the Import PDF dialog
+ * - `-pdfimport` — same import with command-line prompts
+ * - `pdfscale` — calibrate imported PDF geometry from a known distance
+ * - `pdftext` — convert exploded PDF text strokes back into text
+ * - `pdfblock` — turn repeated PDF geometry back into a block
+ * - `pdfdim` — turn exploded PDF dimensions back into dimensions
+ * - `pdfhatch` — turn the loose strokes of a PDF hatch back into a hatch
  */
-export var PDF_PLUGIN_TRIGGERS = ['-cpdf', 'cpdf', 'ipdf'];
+export var PDF_PLUGIN_TRIGGERS = [
+    '-cpdf',
+    'cpdf',
+    'pdfimport',
+    '-pdfimport',
+    'pdfscale',
+    'pdftext',
+    'pdfblock',
+    'pdfdim',
+    'pdfhatch'
+];
+/** Triggers that remain when export is disabled. */
+var PDF_IMPORT_TRIGGERS = [
+    'pdfimport',
+    '-pdfimport',
+    'pdfscale',
+    'pdftext',
+    'pdfblock',
+    'pdfdim',
+    'pdfhatch'
+];
 /**
  * Registers the PDF plugin for lazy loading.
  *
@@ -83,8 +109,7 @@ export function registerLazyPdfPlugin(pluginManager, options) {
     if (options === void 0) { options = {}; }
     var disableExport = options.disableExport === true;
     var triggers = disableExport
-        ? ['ipdf']
-        : __spreadArray([], __read(PDF_PLUGIN_TRIGGERS), false);
+        ? __spreadArray([], __read(PDF_IMPORT_TRIGGERS), false) : __spreadArray([], __read(PDF_PLUGIN_TRIGGERS), false);
     pluginManager.registerLazyPlugin({
         name: PDF_PLUGIN_NAME,
         triggers: __spreadArray([], __read(triggers), false),

package/lib/register.js.map

@@ -1 +1 @@
-{"version":3,"file":"register.js","sourceRoot":"","sources":["../src/register.ts"],"names":[],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAEA,8CAA8C;AAC9C,MAAM,CAAC,IAAM,eAAe,GAAG,WAAW,CAAA;AAE1C;;;;;;;GAOG;AACH,MAAM,CAAC,IAAM,mBAAmB,GAAG,CAAC,OAAO,EAAE,MAAM,EAAE,MAAM,CAAU,CAAA;AAErE;;;;;;;GAOG;AACH,MAAM,UAAU,qBAAqB,CACnC,aAAgC,EAChC,OAAyC;IAF3C,iBAiBC;IAfC,wBAAA,EAAA,YAAyC;IAEzC,IAAM,aAAa,GAAG,OAAO,CAAC,aAAa,KAAK,IAAI,CAAA;IACpD,IAAM,QAAQ,GAAG,aAAa;QAC5B,CAAC,CAAE,CAAC,MAAM,CAAW;QACrB,CAAC,0BAAK,mBAAmB,SAAC,CAAA;IAE5B,aAAa,CAAC,kBAAkB,CAAC;QAC/B,IAAI,EAAE,eAAe;QACrB,QAAQ,2BAAM,QAAQ,SAAC;QACvB,MAAM,EAAE;;;;4BACsB,qBAAM,MAAM,CAAC,2BAA2B,CAAC,EAAA;;wBAA7D,eAAe,GAAK,CAAA,SAAyC,CAAA,gBAA9C;wBACvB,sBAAO,eAAe,CAAC,EAAE,aAAa,eAAA,EAAE,CAAC,EAAA;;;aAC1C;KACF,CAAC,CAAA;AACJ,CAAC"}
+{"version":3,"file":"register.js","sourceRoot":"","sources":["../src/register.ts"],"names":[],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAEA,8CAA8C;AAC9C,MAAM,CAAC,IAAM,eAAe,GAAG,WAAW,CAAA;AAE1C;;;;;;;;;;;;;GAaG;AACH,MAAM,CAAC,IAAM,mBAAmB,GAAG;IACjC,OAAO;IACP,MAAM;IACN,WAAW;IACX,YAAY;IACZ,UAAU;IACV,SAAS;IACT,UAAU;IACV,QAAQ;IACR,UAAU;CACF,CAAA;AAEV,oDAAoD;AACpD,IAAM,mBAAmB,GAAG;IAC1B,WAAW;IACX,YAAY;IACZ,UAAU;IACV,SAAS;IACT,UAAU;IACV,QAAQ;IACR,UAAU;CACF,CAAA;AAEV;;;;;;;GAOG;AACH,MAAM,UAAU,qBAAqB,CACnC,aAAgC,EAChC,OAAyC;IAF3C,iBAiBC;IAfC,wBAAA,EAAA,YAAyC;IAEzC,IAAM,aAAa,GAAG,OAAO,CAAC,aAAa,KAAK,IAAI,CAAA;IACpD,IAAM,QAAQ,GAAG,aAAa;QAC5B,CAAC,0BAAK,mBAAmB,UACzB,CAAC,0BAAK,mBAAmB,SAAC,CAAA;IAE5B,aAAa,CAAC,kBAAkB,CAAC;QAC/B,IAAI,EAAE,eAAe;QACrB,QAAQ,2BAAM,QAAQ,SAAC;QACvB,MAAM,EAAE;;;;4BACsB,qBAAM,MAAM,CAAC,2BAA2B,CAAC,EAAA;;wBAA7D,eAAe,GAAK,CAAA,SAAyC,CAAA,gBAA9C;wBACvB,sBAAO,eAAe,CAAC,EAAE,aAAa,eAAA,EAAE,CAAC,EAAA;;;aAC1C;KACF,CAAC,CAAA;AACJ,CAAC"}