Hi Phil:

I have 2 scenarios here:

Scenario #1 (Works fine).

When I start canned API server on Python >> (Python CLI prompt)

>>> subprocess.Popen(["./startAPIServer"],stdout=subprocess.PIPE)                               
<subprocess.Popen object at 0x7f48c5825b10>

and run the "python add.py" (which does a curl POST) on regular $ prompt, everything works fine and dandy as expected.

$ python add.py
IP is 192.168.1.15
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   165  100   132  100    33   1848    462 --:--:-- --:--:-- --:--:--  1859

Values of Controller IDs and Controller Names that were successfully added.

Scenario #2.(Canned Server crashes and js error)

When I run both startServer script and add.py from python interactive mode,

>>> subprocess.Popen(["./startAPIServer"],stdout=subprocess.PIPE)
<subprocess.Popen object at 0x7f48c5825b10>
>>> subprocess.call("python add.py", shell=True,stderr=subprocess.PIPE)
IP is 192.168.1.15

Values of Controller IDs and Controller Names that were successfully added.

>>> 
events.js:72
        throw er; // Unhandled 'error' event
              ^
Error: write EPIPE
    at errnoException (net.js:901:11)
    at Object.afterWrite (net.js:718:19)

As you have seen, Scenario #2 also works but after 4-5 seconds, I get the error message (which I presume if from Canned) and the Canned server crashes. Would you know why it would crash??

Just to clarify, the "add.py" does this :

proc = subprocess.Popen(["curl -X POST -i -H \"Content-type: application/json\" -X POST http://localhost:8080/v1/controllers -d '{\"url\": \"http://%s:%d\"}'" % (ip, i)], stdout=subprocess.PIPE, shell=True)
(out, error) = proc.communicate()

If canned is used to stub out requests for a server it can be useful to proxy requests for unknown routes. The idea is to instead of serving a 404 forward the request to specified server and serve the result.

$ canned ./example --proxy=http://example.com 

when querying for a path that is not present in canned like

$ curl http://localhost:3000/i_dont_exist_in_canned

it shoudl issue a query to http://example.com/i_dont_exist_in_canned and serve
whatever the result of this query is.

I have a file called _userAccount.get.json living in some/path/any/another/path/_userAccount.get.json

I want this file to be returned as a JSON response when my client accesses http://localhost:3000/some/path/12345/another/path/userAccount but with canned 0.3.7 this doesn't work. Canned logs: request: get /some/path/12345/another/path/userAccount not found

If I move the file to some/path/any/another/path/userAccount/index.get.json and access http://localhost:3000/some/path/12345/another/path/userAccount/ from the client the expected JSON is returned.

So it seems like the any wildcard is working for index.get.json files, but not for _anyFileName.get.json files.

I've noticed that if I use the:

// !params: { "name": "value"}

notation to use a single JSON file to return different responses according to URL parameter value, this will break if the JSON response contains an apostrophe (and maybe some other "special" characters).

Here's an example showing that using the parameter in the filename works, but using the special comment notation does not work.

_test?param=a.get.json

File contains:

{
    "value": "This is value a's response."
}

http://localhost:3000/demo/test?param=a

{
value: "This is value a's response."
}

Works as expected.

_test?param=b.get.json

File contains:

{
    "value": "This is value b's response."
}

http://localhost:3000/demo/test?param=b

{
value: "This is value b's response."
}

Works as expected.

test2.get.json

File contains:

//! params: {"param": "a"}
{
    "value": "This is value a response."
}
//! params: {"param": "b"}
{
    "value": "This is value b response."
}

http://localhost:3000/demo/test2?param=a

{
value: "This is value a response."
}

Works as expected but note I removed the apostrophe from the file.

http://localhost:3000/demo/test2?param=b

{
value: "This is value b response."
}

Works as expected but note I removed the apostrophe from the file.

_test3.get.json

File contains:

//! params: {"param": "a"}
{
    "value": "This is value a's response."
}
//! params: {"param": "b"}
{
    "value": "This is value b's response."
}

http://localhost:3000/demo/test3?param=a

Internal Server error invalid input file

Logs:

request: get /demo/test3?param=aproblem sanatizing content for _test3.get.json SyntaxError: Unexpected end of input served via: ./demo/_test3.get.json

Broken because of ' in response?

http://localhost:3000/demo/test3?param=b

Internal Server error invalid input file

Logs:

request: get /demo/test3?param=bproblem sanatizing content for _test3.get.json SyntaxError: Unexpected end of input served via: ./demo/_test3.get.json

Broken because of ' in response?

I am using node v0.12.0 on Mac OS X 10.10.2. I think this is a problem in the JSON "comment parser", happy to take a look at addressing this, just wanted to see if this is a problem or something in my environment.

Currently canned resolves the routes on request, which is great since they can be modified while canned is running without any problems, adding new files/folders/etc. One thing missing so is the option to actually list what is available, currently this is mostly done by using tree to show the folder structure of the example folder, which prints something like this:

Sadly this does not take into account the rules canned has to resolve the paths, it would be awesome to get something like this:

example(master*)$ canned routes
verb   path
GET  /search
PUT  /comment/1
...

This means crawling the current file tree and resolving the routes and print.

currently running

$ canned

shows nothing util the first request, this should be fixed to show some info about the server starting

It would be nice to specify the directory as well as the port via the command line, so we should add some command line parsing here

node version: v6.1.0
Whenever you try to use query string parameters canned crashed with the following message

if(source.hasOwnProperty(prop)){
TypeError: source.hasOwnProperty is not a function

When I start canned in the spec/test_responses folder and I request a file with query params I get the correct file served but the output in the log shows the wrong file. This is not that bad but really confusing when mocking services which rely heavily on query params.

Include #110 changes
release

I have the following file-structure:

/api
/api/index.get.html
/api/customer/index.get.json

I start canned via $ canned -p 5000 ./api

localhost:5000/ works.
localhost:5000/customer is not working -> request: get /customer/ not found

When I replace /api/customer/index.get.json with /api/customer/index.get.html it works as HTML-file.
I also tested all other possibilities from the docs, but I cannot get json-files to work :/

canned: 0.3.8
node: 5.11.0

I'm using ecryptfs on Ubuntu and when I install canned, the _regexmatchbug file pushes the filename length over the limit.

Since this file is there to cover a regression I'm loathe to get rid of it but I don't know what other approach I can take to mitigate the problem.

I've appended the npm error output for clarity.

Currently I've been installing it into /tmp with npm install canned then removing the regexmatchbug file and then moving the directory to my project node_modules.

npm ERR! tar.unpack untar error /home/peter/.npm/canned/0.3.7/package.tgz
npm WARN enoent ENOENT: no such file or directory, open '/home/peter/projects/package.json'
npm WARN projects No description
npm WARN projects No repository field.
npm WARN projects No README data
npm WARN projects No license field.
npm ERR! Linux 4.2.0-25-generic
npm ERR! argv "/home/peter/.nvm/versions/node/v4.2.4/bin/node" "/home/peter/.nvm/versions/node/v4.2.4/bin/npm" "install" "canned"
npm ERR! node v4.2.4
npm ERR! npm  v3.5.3
npm ERR! path /home/peter/projects/node_modules/.staging/canned-55c23294/spec/test_responses/_regexmatchbug?u=root&p=root&q=select+mean(value)+from+%22Coraid.1Controller.ZFS.VOps-3008.gauge.wlat%22+where+time+%3E++now()+-+86400000000u+and+time+%3C+now()+-+0u+group+by+time(240000000u)+fill(null).get.json
npm ERR! code ENAMETOOLONG
npm ERR! errno -36
npm ERR! syscall open

npm ERR! ENAMETOOLONG: name too long, open '/home/peter/projects/node_modules/.staging/canned-55c23294/spec/test_responses/_regexmatchbug?u=root&p=root&q=select+mean(value)+from+%22Coraid.1Controller.ZFS.VOps-3008.gauge.wlat%22+where+time+%3E++now()+-+86400000000u+and+time+%3C+now()+-+0u+group+by+time(240000000u)+fill(null).get.json'
npm ERR! 
npm ERR! If you need help, you may report this error at:
npm ERR!     <https://github.com/npm/npm/issues>

npm ERR! Please include the following file with any support request:
npm ERR!     /home/peter/projects/npm-debug.log

This could come in handy when troubleshooting within a dev/test environment.
The feature could be configurable, preferably via argument.

If I can get time, I'll see if I can submit a PR.

discovered while working on #58 combining statusCode and headers on seperate lines with //! does not work

//! statusCode: 201
//! body: { "foo": "bar" }

This should be handled by extractOptions which is only handling the status code right now.

I observe the following:

pmu:tmp pmu$ mkdir api
pmu:tmp pmu$ cat > api/posts.json
[ "123" ]
pmu:tmp pmu$ canned
starting canned on port 3000 for ./
request: get /api/posts.json
/usr/local/share/npm/lib/node_modules/canned/lib/response.js:17
  this.cors_enabled = !!options.cors_enabled
                               ^
TypeError: Cannot read property 'cors_enabled' of undefined
    at new Response (/usr/local/share/npm/lib/node_modules/canned/lib/response.js:17:32)
    at /usr/local/share/npm/lib/node_modules/canned/canned.js:194:48
    at Object.oncomplete (fs.js:107:15)

Maybe I am missing something?

Suppose I have a dir called "Abc" and I have 10 sub-directories in it "1" through "10". I want to start the canned API server for each of those - it works only with relative path ../../xxx/x/xx and not with absolute path (./canned -p 3000 ../zy/tz/ex) -> this works because it is relative. Why is that? Can it not work with absolute path?

Hi there,

This looks like a great mocking framework and exactly what I need for testing, so thanks for putting in all the hard work and sharing this! However I am looking to be able to simulate quite slow services with latency etc. So it would be a really great feature to be able to add a comment to set a delay

i.e. for 2000ms delay
//! delay: 2000

I know the obvious answer is to have a look myself and submit a pull request. So I will try if I get a chance, but I am new to node, so not sure if in reality I will or not. So just thought I would raise the suggestion.

Any tips on a starting point to implement this would also be great! :)

Cheers,
Rob

Had this issue when installing canned with yarn 1.0.1 on Windows. I raised an issue over on Yarn and the issue was closed as being to fault with the canned package name.

What is the current behavior?
error An unexpected error occurred: "https://registry.yarnpkg.com/canned/-/canned-0.3.9.tgz: ENOENT: no such file or directory, open 'C:\Users\samjo\AppData\Local\Yarn\cache\v1\npm-canned-0.3.9-b0bfdeffdc595dd5c0055730b808d5c26b4caa02\spec\test_responses\_a?name=Batman&age=30.get.html'".

Steps to reproduce.
Running yarn add or in fact any yarn invocation after adding "canned": "^0.3.10" as a dev dependency to my package.json results in the error. Amending to use other versions of canned, e.g. 0.3.9 produce the same error.

Cause

this is noy Yarn's fault but a problem with the package itself. That file has ? in its name and that character is not allowed in file or folder names on Windows: https://support.microsoft.com/en-us/help/905231/information-about-the-characters-that-you-cannot-use-in-site-names--fo

All of this cited from the issue raised on the yarn github:
yarnpkg/yarn#4427

Be able to configure and run this as a grunt task. Suggestion for future improvement.

Hello,

I am trying to use the "any" wildcard, but in the middle of the path, for example:

http://127.0.0.1:3000/213/pic

My folder hierarchy is:
/any/pic/index.get.json

Thanks,
Omer

Currently the process of updating the documentation page involves

git checkout gh-pages
git checkout master -- README.md
echo "---\nlayout: index\n---" | cat - README.md > index.md
git add -A
git commit -m "update docs"
git push origin gh-pages

this is an error prone process, so It would be nice if we could do

make docs

Canned is working great for simple file names such as v1/info/index.get.json. But when we have complicated names that we are testing from influxdb, it's giving an error saying that can't interpret as regular expression. Please suggest!

curl -sLGw "\n%{http_code} %{url_effective}\n" "http://localhost:17022/db/EtherCloudStats/series?u=root&p=root&q=select+mean(value)+from+%22Coraid.1Controller.ZFS.VOps-3008.gauge.wlat%22+where+time+%3E++now()+-+86400000000u+and+time+%3C+now()+-+0u+group+by+time(240000000u)+fill(null)"

Message from canned:
request: get /db/EtherCloudStats/series?u=root&p=root&q=select+mean(value)+from+%22Coraid.1Controller.ZFS.VOps-3008.gauge.wlat%22+where+time+%3E++now()+-+86400000000u+and+time+%3C+now()+-+0u+group+by+time(240000000u)+fill(null)
/home/msuri/ec/testeng/data/mock/canned/canned.js:29
new RegExp(fname + "(?=." + queryString.split("&").join(")(?=.") + ").+"
^
SyntaxError: Invalid regular expression: /(?=._u=root)(?=._p=root)(?=.*q=select+mean(value)+from+%22Coraid.1Controller.ZFS.VOps-3008.gauge.wlat%22+where+time+%3E++now()+-+86400000000u+and+time+%3C+now()+-+0u+group+by+time(240000000u)+fill(null)).+get/: Nothing to repeat
at new RegExp ()
at matchFileWithExactQuery (/home/msuri/ec/testeng/data/mock/canned/canned.js:29:5)
at getFileFromRequest (/home/msuri/ec/testeng/data/mock/canned/canned.js:51:13)
at Canned._responseForFile (/home/msuri/ec/testeng/data/mock/canned/canned.js:123:20)
at /home/msuri/ec/testeng/data/mock/canned/canned.js:168:10
at Object.oncomplete (fs.js:107:15)

The files don't provide a way to comment the fields or even the whole file, this would ease development and documentation.

something like

// this is a comment
{
  "key":"value",
  // another comment
  "keyTwo":"oneMoreValue"
}

should result in

{ "key":"value", "keyTwo":"oneMoreValue" } 

I would like to be able to add custom HTTP headers on responses so I can fake - for example - the login process (which returns a Authorization header)

Canned breaks if params contains strange characters.
Like: - + %20 etc.

Example:

//! params: {"category": "jonge-vrouwen", "limit": 40, "offset": 0, "withPhotos": 1}

If you rename/delete a folder and make a request to the old location, the following error appears:

/usr/local/share/npm/lib/node_modules/canned/canned.js:19
  for (i = 0, e = files[i]; e != null; e = files[++i]) {
                   ^
TypeError: Cannot read property '0' of undefined
    at scanFileListForName (/usr/local/share/npm/lib/node_modules/canned/canned.js:19:24)
    at responseForFile (/usr/local/share/npm/lib/node_modules/canned/canned.js:54:10)
    at /usr/local/share/npm/lib/node_modules/canned/canned.js:103:9    at Object.oncomplete (fs.js:107:15)

When on windows checking out a repository which contains files with ? or & in their names will fail, since they are not valid chars for windows filenames. This is trouble when using get parameters encoded in the filename.

One way to solve this would be to add the parameters like a special comment as StatusCode already is. In this case their is the need to provide a way to have multiple files for the same path then. This could be solved by prefixing the filename with a number 1_search.get.xml would map to /search then.

Maybe this can be extended to be used for post and put body as well.

The HTTP Accept header is a mechanism for the client to tell the server which media types are acceptable for the response.

Canned only checks the first subtype in the Accept header, but if a file does not exist with the first subtype, it does not attempt to check for a file matching the following subtypes.

Allow to set the http response code, so we can mock errors like 404 with a json response

maybe via file name resource.get.json.404

Hey there. Thanks for canned - it's awesome!
I went to use the response_delay feature locally, but it looks like the latest package on npm doesn't have that code. Guessing that publishing to npm was just forgotten after that change.

Not sure what we need todo here but it would be nice if it could be easily integrated with express during testing / development. Something like

var app = express()
app.configure("development", function() {
  express.use(canned("./"))
})

Would be awesome.

parse error: Invalid numeric literal at line 2, column 9 for invalid json

Hiya,

here we go: canned.js and parser.js

So in a nutshell, what I did:
I turned canned into an express app and exposed a serve method which allows it to be used in express with app.use(canned.serve(myDir)) (obviously in the latter case it won't set up an express server of its own)
Then, parser.js reads the contents of the to-be-serverd directory, strips the extension of the filename and maps the result to the full file path in a hash. Then when a request is made it uses the hash to look up what file it needs to read.
If the file is .js it's required and the exposed function is called. In all other cases the file is read and it's contents are searched for headers:

/*headers
{
    "Content-Type": "application/vnd.custom+json",
    "Content-Disposition": "attachment"
}
*/

Headers must be declared as a json object, it has a number of benefits, but the main one is: you can create a .js file which reads a json file (depending on the route for instance, or the number of times the route is hit) and dumps it as a header in the output.
If the Content-type header is not set, mime is used to determine the content type by file extension.

Routes are resolved as follows: it will always try to find the file that closest matches the entire url first. E.g: GET /foo/bar/baz?qux=something will match the files, in the following order: (the file name is code-formatted below)

• /foo/bar/baz?qux=something/GET.*
• /foo/bar/baz?qux=something.*
• /foo/bar/baz/GET.*
• /foo/bar/baz.*
• /foo/bar/GET.*
• /foo/bar.*
• /foo/GET.*
• /foo.*
• /GET.*

I know it's not backwards compatible, but IMO this is a cleaner and more flexible approach. And, there's not a large user base (yet), so I think we could get away with it. + migration is pretty easy.

It's quite a different approach, just let me know what you think!!

I haven't updated the bin nor the documentation or tests yet, I wanted to hear your opinion first...

Currently the documentation lives in the readme, which is not ideal, maybe consider moving it to something like readthedocs.org for better viewablity and access.

What would be involved in doing this?

How does this work?

I set my json file for multiple responses like below

//! params: {"limit": 40, "offset": 0}
{
  "name": "value1"
}

//! params: {"limit": 40, "offset": 40}
{
  "name": "value2"
}

But canned continue to deliver only the first response, despite I use a different query_list.
Any idea?

I have a RestAPI which I poll from UI to get status of an operation. The API returns inprogress/success/failure for the same request.
I need to mock this API to first show inprogress and then succeeded.

Is there any way I can do this with canned?

"example/comment/.get.josn"
It is misprint ot something like that?

Having multiple custom headers, such as the one from the readme:

//! customHeaders: [{"MyCustomHeaderName": "MyCustomHeaderValue"}, {"SecondHeaderName": "SecondHeaderValue"}] 

returns the error:

Invalid file header format try //! statusCode: 201

I believe this is due to the code on line 130 of canned.js:

var content = line.split(',').map(function (s) {

which splits the line on commas but should not actually be splitting on the comma located in the JSON string for the custom headers.

I have a simple file with a single Json object.

filename: index.get.json

    //! statusCode: 200
    { "content": "I am a wildcard comment for any id", "author": "sideshowcoder" }

After starting canned I open my browser and type http://localhost:3000/test into the address bar.
I see no output in chrome, safari, or firefox, and I get this response from canned:

  request: get /test not found

Chrome shows this error in the developer tools console:

  Failed to load resource: the server responded with a status of 404 (Not Found)

From the terminal I type:

   $ curl http://localhost:3000/test

and get this response:

   {"content":"I am a wildcard comment for any id","author":"sideshowcoder"}

and this output from canned:

  request: get /test served via: ./index.get.json

Scanning through the directory to explore the API is nice but it would be great if it would be possible to output a simple HTML page which just lists the available endpoints with their requests.

Something like seen in the readme for showing what canned can do.

Most API's have a default error message like this:

{"message":"404 Not found"}

It would be awesome if we could create a 404.json or something, and have a default error response from canned. Any thoughts on name and other functionality? I might take a stab at this soon.

On Windows 7. Having a file as follows with CR/LFs at the end of each line causes an "Unexpected token". Converting the EOL to LF works as expected.

Node version is 0.12.2 (x64) on Windows 7

//! params: {"serialkey": "abc"}
{
"errorCode": "ERROR1"
}

//! params: {"serialkey": "12121"}
{
"errorCode": "ERROR2"
}

When I have a DELETE endpoint I want to have an empty response body. Currently, canned assumes an empty response body is an error and returns a 500 status code. If the file is read properly I think an empty file should return a 204 status by default.

If i have one GET endpoint, returning HTML and JSON based on Accepts header, it always delivers the HTML content whatever be the Accepts Header.

As far as i see, if there is HTML and JSON responses for a endpoint, canned should check the Accepts Header to decide which content to sent

I'm trying to figure out how to use the following example from the docs:

/comment/any/votes/index.get.json

it seems to me like requests such as
GET /comment/1/votes/
GET /comment/asdfasdf/votes/
or even
GET /comment/2/votes

should all map to this file.

However, that doesn't appear to be the behavior I'm seeing. Instead, I get 404 errors. Is this the designed behavior or is it a bug?

So when I have the Canned API server running and I make changes to a an "index.get.json" file, I do not have to restart my server (verified that and it works fine). I can continue sending requests and get responses (including the updated data from index.get.json). My question is if I have all the data in a folder called "Abc" and I sym link it "Def" in such a way that any requests that I send to "Abc" should be re-directed to "Def", that does not seem to work unless I re-start the canned API server. How is that handled/supposed to be handled in the source code?

Version from about 2 weeks ago is working. So not sure if some configuration has changed, or if I need to start/configure it differently. These are the steps that I followed:

1. Download version canned-0.3.1.zip, and unzipped it
2. Ran canned for a particular directory
3. Using curl for a response is giving 404
   starting canned on port 17022 for mockInflux/mockInfluxDB2
   request: get /db/EtherCloudStats/series not found
   request: get /db/EtherCloudStats/series?u=root&p=root&q=select+mean.get.json not foundHere is corresponding gist:

https://gist.github.com/mamtasuri-coraid/36d25df8e0c376b9fd91

With the just merged PR #39 canned gained some functions worth pushing up to npm, and creating a new release... DO IT!

• update version
• update changelog
• push to npm (with tag)
• update docs page

• bump version
• update changelog
• tag
• push to npm

@mulderp pointed out rightfully that the getting started is located rather awkward at the middle if the docs. I think to move it to the top would be great as long as their is a small example in there which explains more then just how to start canned but includes a simple one file setup maybe?