tcmd: Fixes regex reuse in inline tokenizer,

tcmd: changes popover syntax to allow for embeddable markdown
This commit is contained in:
2024-02-29 00:02:04 -07:00
parent ff0a4280e2
commit ce83bdf7af
7 changed files with 147 additions and 154 deletions

View File

@@ -1,7 +1,7 @@
import { zipArrays } from "../zip";
import { inlineTokens } from "./inlineTokens";
import { singleLineTokens } from "./singleLineTokens";
import { singleLineTokens, tokenizeLine } from "./tokenizeLine";
import { tokenizeBlock } from "./tokenizeBlock";
import { tokenizeInline } from "./tokenizeInline";
import { tokenizeParagraph } from "./tokenizeParagraph";
export const createElements = (body: string) => {
@@ -11,6 +11,8 @@ export const createElements = (body: string) => {
};
const tokenize = (body: string) => {
body = body.replace(/\n?<!--(.*?)-->\n?/gs, "");
const paragraphs = body.split("\n\n");
const blockTokens: BlockToken[] = [];
@@ -97,129 +99,3 @@ const tokenize = (body: string) => {
return blockTokens.filter((b) => !b.parent);
};
// const __tokenize = (md: string) => {
// const tokens: (Token)[] = [];
// // md = md.replace(/(?<=[a-z])\n(?=[a-z])/g, " ");
// const lines = md.split("\n");
// let preserveEmpty = false;
// let multilineLines;
// let tokenSettings;
// for (let line of lines) {
// if (!line && !preserveEmpty) continue;
// let foundLine = false;
// if (!multilineLines) {
// token:
// for (const token of multilineTokens) {
// if (!token.rx.test(line)) continue token;
// tokenSettings = token;
// multilineLines = token.create(tokens);
// preserveEmpty = true;
// foundLine = true;
// multilineLines.push({
// type: "text",
// line: token.replace(line),
// });
// }
// } else {
// foundLine = true;
// if (tokenSettings?.closeRx?.test(line) || tokenSettings?.rx.test(line)) {
// tokenSettings = undefined;
// multilineLines = undefined;
// preserveEmpty = false;
// } else {
// multilineLines.push({
// type: "text",
// line,
// });
// }
// }
// if (!multilineLines) {
// token:
// for (const token of singleLineTokens) {
// if (!token.rx.test(line)) continue token;
// foundLine = true;
// line = line.replace(token.replaceRx, "").trim();
// const lineContent = tokenizeInline(line);
// token.create(lineContent, tokens);
// }
// }
// if (foundLine) continue;
// tokens.push({
// type: "text",
// line: tokenizeInline(line),
// });
// }
// return tokens;
// };
const tokenizeLine = (
line: string,
previous?: SingleLineToken,
): SingleLineToken => {
for (const token of singleLineTokens) {
if (!token.rx.test(line)) continue;
const t = token.create(line);
if (t.type === "h2") {
}
t.line = tokenizeInline(line.replace(token.replaceRx, ""));
return t;
}
if (previous?.mends) {
previous.raw += " " + line;
previous.line = tokenizeInline(previous.raw.replace(previous.cfg!.rx, ""));
return previous;
}
return {
line: tokenizeInline(line),
type: "text",
raw: line,
};
};
const tokenizeInline = (line: string) => {
line = line.trim();
const originalLine = line;
const insertMarker = "\u{03A9}";
const tokens: InlineTokenInsert[] = [];
for (const token of inlineTokens) {
token.rx.lastIndex = 0;
let match;
while ((match = token.rx.exec(line)) !== null) {
const tokenStart = match.index;
const tokenEnd = match.index + match[0].length;
token.create(match, tokenStart, tokenEnd, tokens);
}
}
if (tokens.length) {
for (const insert of tokens) {
line = line.slice(0, insert.start) +
"".padStart(insert.end - insert.start, insertMarker) +
line.slice(insert.end, line.length);
}
return zipArrays(
line.split(new RegExp(insertMarker + "{2,}")).map((t): InlineToken => ({
content: t,
type: "text",
})),
tokens,
).filter((t) => t.content);
}
return originalLine;
};

View File

@@ -1,3 +1,55 @@
import { zipArrays } from "../zip";
export const tokenizeInline = (line: string, recursive?: boolean) => {
if (recursive) console.log("recursive call");
line = line.trim();
const originalLine = line;
const insertMarker = "\u{03A9}";
const tokens: InlineTokenInsert[] = [];
for (const token of inlineTokens) {
const rx = new RegExp(token.rx);
let match;
while ((match = rx.exec(line)) !== null) {
const tokenStart = match.index;
const tokenEnd = match.index + match[0].length;
const wrappingToken = tokens.find((t) =>
t.start < tokenStart && t.end > tokenStart
);
if (wrappingToken) continue;
let wrappedToken;
while (
(wrappedToken = tokens.findIndex((t) =>
t.start > tokenStart && t.start < tokenEnd
)) !== -1
) {
tokens.splice(wrappedToken, 1);
}
token.create(match, tokenStart, tokenEnd, tokens);
}
}
if (tokens.length) {
for (const insert of tokens) {
line = line.slice(0, insert.start) +
"".padStart(insert.end - insert.start, insertMarker) +
line.slice(insert.end, line.length);
}
return zipArrays(
line.split(new RegExp(insertMarker + "{2,}")).map((t): InlineToken => ({
content: t,
type: "text",
})),
tokens,
).filter((t) => t.content);
}
return originalLine;
};
const joiner = "<><>";
export const inlineTokens: {
rx: RegExp;
@@ -61,16 +113,17 @@ export const inlineTokens: {
},
},
{
rx: /\^\[(.*?)\]\((.*?)\)/g,
rx: /\^\[(.*?)\]<<(.*?)>>/gm,
create(content, start, end, tokens) {
const [_, text, popover] = content;
// tokenizeInline("", true);
tokens.push({
content: text,
end,
start,
type: "popover",
data: {
popover,
popover: tokenizeInline(popover),
},
});
},

View File

@@ -1,3 +1,34 @@
import { tokenizeInline } from "./tokenizeInline";
export const tokenizeLine = (
line: string,
previous?: SingleLineToken,
): SingleLineToken => {
for (const token of singleLineTokens) {
if (!token.rx.test(line)) continue;
const t = token.create(line);
if (t.type === "h2") {
}
t.line = tokenizeInline(line.replace(token.replaceRx, ""));
return t;
}
if (previous?.mends) {
previous.raw += " " + line;
previous.line = tokenizeInline(previous.raw.replace(previous.cfg!.rx, ""));
return previous;
}
return {
line: tokenizeInline(line),
type: "text",
raw: line,
};
};
export const singleLineTokens: SingleLineCfg[] = [
{
rx: /^#\s/,