65 lines
1.4 KiB
TypeScript
65 lines
1.4 KiB
TypeScript
import { describe, expect, it } from "vitest";
|
|
|
|
import {
|
|
aggregatePdfTextLines,
|
|
buildPageContentText,
|
|
normalizePdfText,
|
|
type PdfTextItemSnapshot,
|
|
} from "../src/index.js";
|
|
|
|
function item(
|
|
text: string,
|
|
x: number,
|
|
y: number,
|
|
width = 20,
|
|
height = 10,
|
|
): PdfTextItemSnapshot {
|
|
return {
|
|
text,
|
|
normalizedText: normalizePdfText(text),
|
|
bounds: { x, y, width, height },
|
|
baselineY: y + height,
|
|
hasEol: false,
|
|
};
|
|
}
|
|
|
|
describe("PDF 文本行聚合", () => {
|
|
it("按坐标聚合中文文本并规范化兼容字符", () => {
|
|
const lines = aggregatePdfTextLines(
|
|
[
|
|
item("ＡＢＣ", 10, 100, 30),
|
|
item("中文", 41, 100, 20),
|
|
item("第二行", 10, 120, 40),
|
|
],
|
|
842,
|
|
);
|
|
expect(lines.map((line) => line.normalizedText)).toEqual([
|
|
"ABC中文",
|
|
"第二行",
|
|
]);
|
|
});
|
|
|
|
it("只在页眉页脚坐标带识别独立页码", () => {
|
|
const lines = aggregatePdfTextLines(
|
|
[
|
|
item("章节 1", 72, 400, 50),
|
|
item("— 1 —", 280, 731, 35, 14),
|
|
],
|
|
842,
|
|
);
|
|
expect(lines.map((line) => line.role)).toEqual([
|
|
"content",
|
|
"page-number",
|
|
]);
|
|
expect(buildPageContentText(lines)).toBe("章节 1");
|
|
});
|
|
|
|
it("不会把页脚中的普通说明误判为页码", () => {
|
|
const lines = aggregatePdfTextLines(
|
|
[item("内部资料 1", 72, 800, 70)],
|
|
842,
|
|
);
|
|
expect(lines[0]?.role).toBe("content");
|
|
});
|
|
});
|