(#3877) - refactor binary #3877
Conversation
| @@ -52,8 +53,8 @@ function fetchRequest(options, callback) { | |||
| } | |||
|
|
|||
| if (options.body && (options.body instanceof Blob)) { | |||
| utils.readAsBinaryString(options.body, function(binary) { | |||
| fetchOptions.body = utils.fixBinary(binary); | |||
There was a problem hiding this comment.
Here's where we can simplify by just directly using readAsArrayBuffer.
|
As for file naming conventions, I think we should start moving away from |
|
Another note: this does increase the min+gz size from 47655 to 47725, probably due to the extra browserify boilerplate around each little function. |
| testUtils.atob = function (arg) { | ||
| return PouchDB.utils.atob(arg); | ||
| }; | ||
|
|
There was a problem hiding this comment.
Trying to reduce our dependence on PouchDB.utils in the tests.
|
forgot the commit message format. we can fix on the way in |
Move them out of `utils.js` and into their own separate files. None of this is a functional change, except for a small optimization I noticed in `request-browser.js`, where we can directly use `readAsArrayBuffer` instead of first converting to a binary string.
Move them out of `utils.js` and into their own separate files. None of this is a functional change, except for a small optimization I noticed in `request-browser.js`, where we can directly use `readAsArrayBuffer` instead of first converting to a binary string.
Move them out of `utils.js` and into their own separate files. None of this is a functional change, except for a small optimization I noticed in `request-browser.js`, where we can directly use `readAsArrayBuffer` instead of first converting to a binary string.
|
all makes sense, cheers - 2376abe |
Move them out of
utils.jsand into their own separate files.None of this is a functional change, except for a small optimization
I noticed in
request-browser.js, where we can directly usereadAsArrayBufferinstead of first converting to a binary string.