EbookFoundation
/
free-programming-books-parser
mirror of https://github.com/EbookFoundation/free-programming-books-parser.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

i *think* this parses notes containing boldfaced text

pull/3/head
Paul Kelly III 2021-11-19 02:09:00 -05:00
parent 8b07ef46a2
commit 2e8f4921ce
1 changed files with 54 additions and 29 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

83
index.js
View File

@ -29,45 +29,70 @@ const excludes = [
* @return {Object} Returns an Object containing details about the piece of media Exact format TBD. * @return {Object} Returns an Object containing details about the piece of media Exact format TBD.
*/ */
let parseListItem = function (listItem) { let parseListItem = function (listItem) {
let stripParens = function (s) {
if (s.slice(0,1) === "(" && s.slice(-1) === ")")
return s.slice(1,-1);
return s;
}
let entry = {}; let entry = {};
let s = ""; // If we need to build up a string over multiple listItem elements
const [link, ...otherStuff] = listItem; // head of listItem = url, the rest is "other stuff" const [link, ...otherStuff] = listItem; // head of listItem = url, the rest is "other stuff"
entry.url = link.url; entry.url = link.url;
entry.title = link.children[0].value; entry.title = link.children[0].value;
// remember to get OTHER STUFF!! remember there may be multiple links! // remember to get OTHER STUFF!! remember there may be multiple links!
for (let i of otherStuff) { for (let i of otherStuff) {
if (i.type === "text" && i.value.slice(0,3) === " - ") { if (s === "") { // this is almost always, except for when we are parsing a multi-element note
// author found if (i.type === "text" && i.value.slice(0,3) === " - ") {
let parenIndex = i.value.indexOf("("); // author found
if (parenIndex === -1) { let parenIndex = i.value.indexOf("(");
entry.author = i.value.slice(3).trim(); if (parenIndex === -1) {
entry.author = i.value.slice(3).trim();
}
else {
entry.author = i.value.slice(3, parenIndex).trim(); // go from " - " until the first "("
}
} }
else { if (i.type === "emphasis" &&
entry.author = i.value.slice(3, parenIndex).trim(); // go from " - " until the first "(" i.children[0].value.slice(0, 1) === "(" && i.children[0].value.slice(-1) === ")") {
// access notes found (currently assumes exactly one child, so far this is always the case)
entry.accessNotes = i.children[0].value.slice(1, -1);
} }
} if (i.type === "link") {
if (i.type === "emphasis" && // other links found
i.children[0].value.slice(0, 1) === "(" && i.children[0].value.slice(-1) === ")") { if (entry.otherLinks === undefined) entry.otherLinks = [];
// access notes found (currently assumes exactly one child, so far this is always the case) entry.otherLinks.push({title: stripParens(i.children[0].value), url: i.url});
entry.accessNotes = i.children[0].value.slice(1, -1); // entry.otherLinks = [...entry.otherLinks, {title: i.children[0].value, url: i.url}]; // <-- i wish i could get this syntax to work with arrays
} }
if (i.type === "link") { if (i.type === "text" && i.value.indexOf("(") !== -1 && i.value.indexOf(")") !== -1) {
// other links found // notes found (currently assumes no nested parentheses)
if (entry.otherLinks === undefined) entry.otherLinks = []; if (entry.notes === undefined) entry.notes = [];
entry.otherLinks.push({title: i.children[0].value, url: i.url}); let leftParen = i.value.indexOf("(");
// entry.otherLinks = [...entry.otherLinks, {title: i.children[0].value, url: i.url}]; // <-- i wish i could get this syntax to work with arrays let rightParen = i.value.indexOf(")", leftParen);
} if (rightParen === -1) {
if (i.type === "text" && i.value.indexOf("(") !== -1) { // there must be some *emphasis* found
// notes found (currently assumes no nested parentheses) s += i.value.slice(leftParen + 1);
if (entry.notes === undefined) entry.notes = []; } else {
let leftParen = i.value.indexOf("("); entry.notes.push(i.value.slice(leftParen + 1, rightParen));
let rightParen = i.value.indexOf(")", leftParen); }
if (rightParen === -1) { // also TODO: if theres more than one disjoint set of parens
// there must be some *emphasis* found }
// TODO: this } else { // for now we assume that all previous ifs are mutually exclusive with this, may polish later
if (i.type === "emphasis") {
// this is the emphasis, add it in boldface and move on
s += "*" + i.children[0].value + "*";
} else { } else {
entry.notes.push(i.value.slice(leftParen + 1, rightParen)); // hopefully this is the end of the note
let rightParen = i.value.indexOf(")");
if (rightParen === -1) {
// we have to go AGAIN
s += i.value;
} else {
// finally, we have reached the end of the note
entry.notes.push(s + i.value.slice(0, rightParen));
s = "";
// TODO: there can be a normal note following a bold-containing note
}
} }
// also TODO: if theres more than one disjoint set of parens
} }
} }
return entry; return entry;