Merge pull request #32857 from overleaf/ds-pandoc-import-md

[WEB + CLSI] Import markdown files using pandoc

GitOrigin-RevId: adad7831ddb13a8fcb8063871166bde13cbbf1b6
This commit is contained in:
Mathias Jakobsen
2026-05-08 08:09:02 +00:00
committed by Copybot
parent 44efc9d745
commit eddcc5a42e
26 changed files with 813 additions and 312 deletions
@@ -36,7 +36,6 @@ describe('ConversionManager', function () {
}
ctx.conversionId = 'test-conversion-id'
ctx.inputPath = '/path/to/input.docx'
ctx.conversionDir = '/compiles/test-conversion-id'
ctx.outputPath = '/compiles/test-conversion-id/output-uuid.zip'
@@ -65,188 +64,287 @@ describe('ConversionManager', function () {
ctx.uuidStub.restore()
})
describe('convertDocxToLaTeXWithLock', function () {
describe('general behavior', function () {
beforeEach(async function (ctx) {
ctx.result =
await ctx.ConversionManager.promises.convertDocxToLaTeXWithLock(
ctx.conversionId,
ctx.inputPath
describe('convertToLaTeXWithLock', function () {
describe('with conversionType=docx', function () {
beforeEach(function (ctx) {
ctx.inputPath = '/path/to/input.docx'
})
describe('file setup and pandoc args', function () {
beforeEach(async function (ctx) {
ctx.result =
await ctx.ConversionManager.promises.convertToLaTeXWithLock(
ctx.conversionId,
ctx.inputPath,
'docx'
)
})
it('should acquire a lock', async function (ctx) {
sinon.assert.calledWith(ctx.LockManager.acquire, ctx.conversionDir)
})
it('should copy the input file to the conversion directory with docx filename', async function (ctx) {
sinon.assert.calledWith(ctx.fs.mkdir, ctx.conversionDir, {
recursive: true,
})
sinon.assert.calledWith(
ctx.fs.copyFile,
ctx.inputPath,
Path.join(ctx.conversionDir, 'input.docx')
)
})
it('should acquire a lock', async function (ctx) {
sinon.assert.calledWith(ctx.LockManager.acquire, ctx.conversionDir)
})
it('should copy the input file to the conversion directory', async function (ctx) {
sinon.assert.calledWith(ctx.fs.mkdir, ctx.conversionDir, {
recursive: true,
})
sinon.assert.calledWith(
ctx.fs.copyFile,
ctx.inputPath,
Path.join(ctx.conversionDir, 'input.docx')
)
})
it('should convert conversion timeout to milliseconds', async function (ctx) {
expect(ctx.CommandRunner.promises.run.firstCall.args[4]).toBe(60_000)
expect(ctx.CommandRunner.promises.run.secondCall.args[4]).toBe(60_000)
})
it('should run pandoc followed by zip in the conversion directory', function (ctx) {
expect(ctx.CommandRunner.promises.run.callCount).toBe(2)
expect(ctx.CommandRunner.promises.run.firstCall.args).toEqual([
ctx.conversionId,
[
'pandoc',
'input.docx',
'--output',
'main.tex',
'--extract-media=.',
'--from',
'docx+citations',
'--to',
'latex',
'--citeproc',
'--standalone',
],
ctx.conversionDir,
ctx.Settings.pandocImage,
60_000,
{},
'conversions',
])
expect(ctx.CommandRunner.promises.run.secondCall.args).toEqual([
ctx.conversionId,
['zip', '-r', 'output-uuid.zip', '.'],
ctx.conversionDir,
ctx.Settings.pandocImage,
60_000,
{},
'conversions',
])
})
})
describe('successful conversion', function () {
beforeEach(async function (ctx) {
ctx.CommandRunner.promises.run.resolves({
stdout: 'mock-stdout',
stderr: 'mock-stderr',
exitCode: 0,
})
ctx.result =
await ctx.ConversionManager.promises.convertDocxToLaTeXWithLock(
it('should convert conversion timeout to milliseconds', async function (ctx) {
expect(ctx.CommandRunner.promises.run.firstCall.args[4]).toBe(60_000)
expect(ctx.CommandRunner.promises.run.secondCall.args[4]).toBe(60_000)
})
it('should run pandoc with docx args followed by zip', function (ctx) {
expect(ctx.CommandRunner.promises.run.callCount).toBe(2)
expect(ctx.CommandRunner.promises.run.firstCall.args).toEqual([
ctx.conversionId,
ctx.inputPath
)
})
it('should remove the source document after conversion', async function (ctx) {
sinon.assert.calledWith(
ctx.fs.unlink,
Path.join(ctx.conversionDir, 'input.docx')
)
})
it('should return the conversion directory', function (ctx) {
expect(ctx.result).toBe(ctx.outputPath)
})
it('should release the lock', function (ctx) {
sinon.assert.called(ctx.lock.release)
})
})
describe('unsuccessful conversion (exitcode)', function () {
beforeEach(async function (ctx) {
ctx.CommandRunner.promises.run.resolves({
stdout: 'mock-stdout',
stderr: 'mock-stderr',
exitCode: 63,
})
await expect(
ctx.ConversionManager.promises.convertDocxToLaTeXWithLock(
[
'pandoc',
'input.docx',
'--output',
'main.tex',
'--to',
'latex',
'--standalone',
'--extract-media=.',
'--from',
'docx+citations',
'--citeproc',
],
ctx.conversionDir,
ctx.Settings.pandocImage,
60_000,
{},
'conversions',
])
expect(ctx.CommandRunner.promises.run.secondCall.args).toEqual([
ctx.conversionId,
ctx.inputPath
)
).to.be.rejectedWith('pandoc conversion failed')
})
it('should remove the entire conversion directory', async function (ctx) {
sinon.assert.calledWith(ctx.fs.rm, ctx.conversionDir, {
force: true,
recursive: true,
['zip', '-r', 'output-uuid.zip', '.'],
ctx.conversionDir,
ctx.Settings.pandocImage,
60_000,
{},
'conversions',
])
})
})
it('should release the lock', function (ctx) {
sinon.assert.called(ctx.lock.release)
})
})
describe('unsuccessful compression (exitcode)', function () {
beforeEach(async function (ctx) {
ctx.CommandRunner.promises.run
.onFirstCall()
.resolves({
stdout: 'mock-pandoc-stdout',
stderr: 'mock-pandoc-stderr',
describe('successful conversion', function () {
beforeEach(async function (ctx) {
ctx.CommandRunner.promises.run.resolves({
stdout: 'mock-stdout',
stderr: 'mock-stderr',
exitCode: 0,
})
.onSecondCall()
.resolves({
stdout: 'mock-zip-stdout',
stderr: 'mock-zip-stderr',
exitCode: 12,
})
await expect(
ctx.ConversionManager.promises.convertDocxToLaTeXWithLock(
ctx.conversionId,
ctx.inputPath
ctx.result =
await ctx.ConversionManager.promises.convertToLaTeXWithLock(
ctx.conversionId,
ctx.inputPath,
'docx'
)
})
it('should remove the source document after conversion', async function (ctx) {
sinon.assert.calledWith(
ctx.fs.unlink,
Path.join(ctx.conversionDir, 'input.docx')
)
).to.be.rejectedWith('pandoc conversion failed')
})
})
it('should remove the entire conversion directory', async function (ctx) {
sinon.assert.calledWith(ctx.fs.rm, ctx.conversionDir, {
force: true,
recursive: true,
it('should return the output zip path', function (ctx) {
expect(ctx.result).toBe(ctx.outputPath)
})
it('should release the lock', function (ctx) {
sinon.assert.called(ctx.lock.release)
})
})
it('should release the lock', function (ctx) {
sinon.assert.called(ctx.lock.release)
describe('unsuccessful conversion (exitcode)', function () {
beforeEach(async function (ctx) {
ctx.CommandRunner.promises.run.resolves({
stdout: 'mock-stdout',
stderr: 'mock-stderr',
exitCode: 63,
})
await expect(
ctx.ConversionManager.promises.convertToLaTeXWithLock(
ctx.conversionId,
ctx.inputPath,
'docx'
)
).to.be.rejectedWith('pandoc conversion failed')
})
it('should remove the entire conversion directory', async function (ctx) {
sinon.assert.calledWith(ctx.fs.rm, ctx.conversionDir, {
force: true,
recursive: true,
})
})
it('should release the lock', function (ctx) {
sinon.assert.called(ctx.lock.release)
})
})
describe('unsuccessful compression (exitcode)', function () {
beforeEach(async function (ctx) {
ctx.CommandRunner.promises.run
.onFirstCall()
.resolves({
stdout: 'mock-pandoc-stdout',
stderr: 'mock-pandoc-stderr',
exitCode: 0,
})
.onSecondCall()
.resolves({
stdout: 'mock-zip-stdout',
stderr: 'mock-zip-stderr',
exitCode: 12,
})
await expect(
ctx.ConversionManager.promises.convertToLaTeXWithLock(
ctx.conversionId,
ctx.inputPath,
'docx'
)
).to.be.rejectedWith('pandoc conversion failed')
})
it('should remove the entire conversion directory', async function (ctx) {
sinon.assert.calledWith(ctx.fs.rm, ctx.conversionDir, {
force: true,
recursive: true,
})
})
it('should release the lock', function (ctx) {
sinon.assert.called(ctx.lock.release)
})
})
describe('unsuccessful conversion (throws)', function () {
beforeEach(async function (ctx) {
ctx.CommandRunner.promises.run.rejects(
new Error('mock conversion error')
)
await expect(
ctx.ConversionManager.promises.convertToLaTeXWithLock(
ctx.conversionId,
ctx.inputPath,
'docx'
)
).to.be.rejectedWith('pandoc conversion failed')
})
it('should remove the entire conversion directory', async function (ctx) {
sinon.assert.calledWith(ctx.fs.rm, ctx.conversionDir, {
force: true,
recursive: true,
})
})
it('should release the lock', function (ctx) {
sinon.assert.called(ctx.lock.release)
})
})
})
describe('unsuccessful conversion (throws)', function () {
beforeEach(async function (ctx) {
ctx.CommandRunner.promises.run.rejects(
new Error('mock conversion error')
)
await expect(
ctx.ConversionManager.promises.convertDocxToLaTeXWithLock(
ctx.conversionId,
ctx.inputPath
)
).to.be.rejectedWith('pandoc conversion failed')
describe('with conversionType=markdown', function () {
beforeEach(function (ctx) {
ctx.inputPath = '/path/to/input.md'
})
it('should remove the entire conversion directory', async function (ctx) {
sinon.assert.calledWith(ctx.fs.rm, ctx.conversionDir, {
force: true,
recursive: true,
describe('file setup and pandoc args', function () {
beforeEach(async function (ctx) {
ctx.result =
await ctx.ConversionManager.promises.convertToLaTeXWithLock(
ctx.conversionId,
ctx.inputPath,
'markdown'
)
})
it('should copy the input file to the conversion directory with md filename', async function (ctx) {
sinon.assert.calledWith(ctx.fs.mkdir, ctx.conversionDir, {
recursive: true,
})
sinon.assert.calledWith(
ctx.fs.copyFile,
ctx.inputPath,
Path.join(ctx.conversionDir, 'input.md')
)
})
it('should run pandoc with markdown args followed by zip', function (ctx) {
expect(ctx.CommandRunner.promises.run.callCount).toBe(2)
expect(ctx.CommandRunner.promises.run.firstCall.args).toEqual([
ctx.conversionId,
[
'pandoc',
'input.md',
'--output',
'main.tex',
'--to',
'latex',
'--standalone',
'--from',
'markdown',
],
ctx.conversionDir,
ctx.Settings.pandocImage,
60_000,
{},
'conversions',
])
expect(ctx.CommandRunner.promises.run.secondCall.args).toEqual([
ctx.conversionId,
['zip', '-r', 'output-uuid.zip', '.'],
ctx.conversionDir,
ctx.Settings.pandocImage,
60_000,
{},
'conversions',
])
})
})
it('should release the lock', function (ctx) {
sinon.assert.called(ctx.lock.release)
describe('successful conversion', function () {
beforeEach(async function (ctx) {
ctx.CommandRunner.promises.run.resolves({
stdout: 'mock-stdout',
stderr: 'mock-stderr',
exitCode: 0,
})
ctx.result =
await ctx.ConversionManager.promises.convertToLaTeXWithLock(
ctx.conversionId,
ctx.inputPath,
'markdown'
)
})
it('should remove the source document after conversion', async function (ctx) {
sinon.assert.calledWith(
ctx.fs.unlink,
Path.join(ctx.conversionDir, 'input.md')
)
})
it('should return the output zip path', function (ctx) {
expect(ctx.result).toBe(ctx.outputPath)
})
})
})
})