@knaw-huc/text-annotation-segmenter 0.3.0 → 0.4.0

package/README.md

@@ -35,13 +35,14 @@ annotation bc:   __
 ```
 
 ```ts
-import { segment } from "@knaw-huc/text-annotation-segmenter";
+import {segment} from "@knaw-huc/text-annotation-segmenter";
 
 const text = 'abc';
 const ab = {id: 'ab', begin: 0, end: 2};
 const bc = {id: 'bc', begin: 1, end: 3};
 
-const segments = segment(text, [ab, bc], (a) => a);
+const getOffsets = annotation => annotation;
+const segments = segment(text, [ab, bc], getOffsets);
 
 expect(segments).toEqual([
   {index: 0, begin: 0, end: 1, body: 'a', annotations: [ab]},
@@ -69,7 +70,8 @@ const text = 'aabb';
 const section = {id: 'section', type: 'section', begin: 0, end: 4};
 const paragraph = {id: 'paragraph', type: 'paragraph', begin: 0, end: 2};
 
-const segments = segment(text, [section, paragraph], (a) => a);
+const getOffsets = annotation => annotation;
+const segments = segment(text, [section, paragraph], getOffsets);
 
 const isGroup = a => a.type === 'section' || a.type === 'paragraph';
 const getId = a => a.id;

package/dist/GetMarkerPosition.d.ts

@@ -0,0 +1,2 @@
+import { MarkerPosition } from './Model.ts';
+export type GetMarkerPosition<T> = (a: T) => MarkerPosition;

package/dist/Model.d.ts

@@ -14,3 +14,4 @@ export type TextSegment<T> = Offsets & {
     annotations: T[];
 };
 export type SegmentIndex = number;
+export type MarkerPosition = 'prefix' | 'postfix';

package/dist/index.d.ts

@@ -4,5 +4,6 @@ export { findSegmentOffsets } from './findSegmentOffsets.ts';
 export type { SegmentOffsets } from './findSegmentOffsets.ts';
 export { groupSegments } from './groupSegments.ts';
 export type { Group, SegmentGroup } from './groupSegments.ts';
-export type { Offsets, TextSegment, SegmentIndex, } from './Model.ts';
+export type { Offsets, TextSegment, SegmentIndex, MarkerPosition, } from './Model.ts';
 export type { GetOffsets } from './GetOffsets.ts';
+export type { GetMarkerPosition } from './GetMarkerPosition.ts';

package/dist/index.js

@@ -1,10 +1,16 @@
-function segment(text, annotations, getOffsets) {
+function segment(text, annotations, getOffsets, getMarkerPosition = () => "postfix") {
   const segments = [];
   let segmentCounter = -1;
   const offsetMap = /* @__PURE__ */ new Map(), getOrCreateOffset = (charIndex) => {
     let offset = offsetMap.get(charIndex);
     return offset || (offset = { charIndex, starting: [], ending: [] }, offsetMap.set(charIndex, offset)), offset;
-  };
+  }, createMarkerSegment = (markers, charIndex) => ({
+    index: ++segmentCounter,
+    begin: charIndex,
+    end: charIndex,
+    body: "",
+    annotations: [...markers, ...activeAnnotations]
+  });
   getOrCreateOffset(0), getOrCreateOffset(text.length);
   for (const annotation of annotations) {
     const offsets = getOffsets(annotation), isMarker = offsets.begin === offsets.end;
@@ -29,23 +35,15 @@ function segment(text, annotations, getOffsets) {
       });
     }
     lastOffset = offset.charIndex;
-    let hasMarkers = !1;
-    const markerSegmentAnnotations = [];
-    for (const a of offset.starting)
-      offset.ending.includes(a) ? (hasMarkers = !0, markerSegmentAnnotations.push(a)) : activeAnnotations.add(a);
+    const postfixMarkers = [], prefixMarkers = [];
+    for (const annotation of offset.starting)
+      offset.ending.includes(annotation) && (getMarkerPosition(annotation) === "prefix" ? prefixMarkers.push(annotation) : postfixMarkers.push(annotation));
+    postfixMarkers.length && segments.push(createMarkerSegment(postfixMarkers, offset.charIndex));
+    for (const annotation of offset.starting)
+      offset.ending.includes(annotation) || activeAnnotations.add(annotation);
     for (const annotation of offset.ending)
       activeAnnotations.delete(annotation);
-    if (hasMarkers) {
-      markerSegmentAnnotations.push(...activeAnnotations);
-      const index = ++segmentCounter;
-      segments.push({
-        index,
-        begin: offset.charIndex,
-        end: offset.charIndex,
-        body: "",
-        annotations: markerSegmentAnnotations
-      });
-    }
+    prefixMarkers.length && segments.push(createMarkerSegment(prefixMarkers, offset.charIndex));
   }
   return segments;
 }

package/dist/segment.d.ts

@@ -1,9 +1,10 @@
 import { TextSegment } from './Model';
 import { GetOffsets } from './GetOffsets.ts';
+import { GetMarkerPosition } from './GetMarkerPosition.ts';
 /**
  * Split a text into {@link TextSegment}s with character offsets and a list of applying annotations.
  */
-export declare function segment<T>(text: string, annotations: T[], getOffsets: GetOffsets<T>): TextSegment<T>[];
+export declare function segment<T>(text: string, annotations: T[], getOffsets: GetOffsets<T>, getMarkerPosition?: GetMarkerPosition<T>): TextSegment<T>[];
 export type AnnotationSegmentsByChar<T> = {
     charIndex: number;
     starting: T[];

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@knaw-huc/text-annotation-segmenter",
-  "version": "0.3.0",
+  "version": "0.4.0",
   "repository": {
     "type": "git",
     "url": "https://github.com/knaw-huc/text-annotation-segmenter.git"