_ = require('underscore')
fs = require('fs')
path = require 'path'
QuotedHTMLTransformer = require('../src/services/quoted-html-transformer')
describe "QuotedHTMLTransformer", ->
readFile = (fname) ->
emailPath = path.resolve(__dirname, 'fixtures', 'emails', fname)
return fs.readFileSync(emailPath, 'utf8')
hideQuotedHTML = (fname) ->
return QuotedHTMLTransformer.hideQuotedHTML(readFile(fname))
removeQuotedHTML = (fname, opts={}) ->
return QuotedHTMLTransformer.removeQuotedHTML(readFile(fname), opts)
numQuotes = (html) ->
re = new RegExp(QuotedHTMLTransformer.annotationClass, 'g')
html.match(re)?.length ? 0
[1..18].forEach (n) ->
it "properly parses email_#{n}", ->
opts = keepIfWholeBodyIsQuote: true
expect(removeQuotedHTML("email_#{n}.html", opts)).toEqual readFile("email_#{n}_stripped.html")
describe 'manual quote detection tests', ->
clean = (str) ->
str.replace(/[\n\r]/g, "").replace(/\s{2,}/g, " ")
# The key is the inHTML. The value is the outHTML
tests = []
# Test 1
tests.push
before: """
Some text
More text
Parent
Sub
Sub Sub
Sub
Text at end
The last quote!
Some text
More text
Parent
Sub
Sub Sub
Sub
Text at end
Yo
Nothing but quotes
"""
after: """
Yo
"""
# Test 3: It found the blockquote in another div
tests.push
before: """
Hello World
"""
after: """
Hello World
"""
# Test 4: It works inside of a wrapped div
tests.push
before: """
"""
after: """
"""
# Test 5: Inline quotes and text
tests.push
before: """
Hello
Inline quote
World
"""
after: """
Hello
Inline quote
World
"""
# Test 6: No quoted elements at all
tests.push
before: """
Hello World
"""
after: """
Hello World
"""
# Test 7: Common ancestor is a quoted node
tests.push
before: """
Content
Some content
More content
Other content
"""
after: """
Content
"""
# Test 8: All of our quote blocks we want to remove are at the end…
# sortof… but nested in a bunch of stuff
#
# Note that "content" is burried deep in the middle of a div
tests.push
before: """
Content
Some content
More content
Other content
Some text quote
Some text
Another quote
More quotes!
"""
after: """
Content
Some content
More content
Other content
Some text quote
Some text
More text
"""
# Test 9: Last several tags are blockquotes. Note the 3 blockquote
# at the end, the interstital div, and the blockquote inside of the
# first div
tests.push
before: """
I'm inline
Content
Remove me
Foo
Bar
Baz
"""
after: """
"""
# Test 10: If it's only a quote and no other text, then just show the
# quote
tests.push
before: """
Nothing but quotes
"""
after: """
Nothing but quotes
"""
# Test 11: The tag itself is just a quoted text block.
# I believe this is https://sentry.nylas.com/sentry/edgehill/group/8323/
tests.push
before: """
This entire thing is quoted text!
"""
after: """
"""
# Test 12: Make sure that a single quote inside of a bunch of other
# content is detected. We used to have a bug where we were only
# looking at the common ancestor of blockquotes (and if there's 1 then
# the ancestor is itself). We now look at the root document for
# trailing text.
tests.push
before: """
Yo
Yo
"""
after: """
Yo
Yo
"""
# Test 13: If there's an "On date…" string immediatley before a blockquote,
# then remove it.
tests.push
before: """
Hey
On FOOBAR
On Thu, Mar 3, 2016
at 3:19 AM,
First Middle Last-Last
<
test@nylas.com
>
wrote:
QUOTED TEXT
"""
after: """
Hey
On FOOBAR
"""
# Test 14: Don't pick up false positives on the string precursors to block
# quotes.
tests.push
before: """
Hey
On FOOBAR
On Thu, Mar 3, 2016 I went to my writing club and wrote:
A little song
QUOTED TEXT
On FOOBAR
On Thu, Mar 3, 2016 I went to my writing club and wrote:
A little song
"""
it 'works with these manual test cases', ->
for {before, after} in tests
opts = keepIfWholeBodyIsQuote: true
test = clean(QuotedHTMLTransformer.removeQuotedHTML(before, opts))
expect(test).toEqual clean(after)
it 'removes all trailing
tags except one', ->
input0 = "hello world
foolololol
"
expect0 = "hello world
"
expect(QuotedHTMLTransformer.removeQuotedHTML(input0)).toEqual expect0
it 'preserves
tags in the middle and only chops off tail', ->
input0 = "hello
world
foolololol
"
expect0 = "hello
world
"
expect(QuotedHTMLTransformer.removeQuotedHTML(input0)).toEqual expect0
it 'works as expected when body tag inside the html', ->
input0 = """
On Dec 16 2015, at 7:08 pm, Juan Tejada <juan@nylas.com> wrote:
h2
he he hehehehehehe
dufjcasc
"""
expect0 = "
"
expect(QuotedHTMLTransformer.removeQuotedHTML(input0)).toEqual expect0
# We have a little utility method that you can manually uncomment to
# generate what the current iteration of the QuotedHTMLTransformer things the
# `removeQuotedHTML` should look like. These can be manually inspected in
# a browser before getting their filename changed to
# `email_#{n}_stripped.html". The actually tests will run the current
# iteration of the `removeQuotedHTML` against these files to catch if
# anything has changed in the parser.
#
# It's inside of the specs here instaed of its own script because the
# `QuotedHTMLTransformer` needs Electron booted up in order to work because
# of the DOMParser.
xit "Run this simple funciton to generate output files", ->
[18].forEach (n) ->
newHTML = QuotedHTMLTransformer.removeQuotedHTML(readFile("email_#{n}.html"))
outPath = path.resolve(__dirname, 'fixtures', 'emails', "email_#{n}_raw_stripped.html")
fs.writeFileSync(outPath, newHTML)