fix(api): Remove crs from scripts in api/resources

If a wheel is built on Windows, there's a good chance there will be carriage returns in the scripts
in api/resources, which break the container initialization scripts. To avoid editing the files in
question in the repo, add a step during the wheel build that unzips the wheel, removes carriage
returns, and rezips it.

Closes #2305

CONTRIBUTING.md

@@ -126,6 +126,7 @@ Individual projects may have additional instructions, so be sure to check out th
 Your computer will need the following tools installed to be able to develop with the Opentrons platform:
 
 *   macOS 10.11+, Linux, or Windows 10
+    *   On Windows, please configure your git client to `autocrlf=input` if you intend to make API server changes and push them to the robot.
 *   Python 3.6 ([pyenv](https://github.com/pyenv/pyenv) is optional, but recommended for macOS / Linux. If `pyenv` is not available for your system or you do not want to use it, you can set the environment variable `OT_PYTHON` to the full path to the Python 3.6 executable)
 
     ```shell

api/Makefile

@@ -61,6 +61,7 @@ dev:
 .PHONY: wheel
 wheel: clean
 	$(python) setup.py bdist_wheel
+	$(python) remove_cr_in_zip.py -v dist/
 	shx rm -rf build
 	shx ls dist
 

api/remove_cr_in_zip.py

@@ -0,0 +1,118 @@
+"""
+Edit an opentrons wheel to remove any carriage returns from files in resources
+
+If there are crs in the scripts in resources/, the robot will not be able to
+boot because those files are sourced directly during the boot process. While
+the crs may be rewritten when they are committed and pushed to github, if a
+windows dev does a push-api from their local checkout there's a good chance
+the crs will still be present.
+
+This script therefore opens the specified wheel, unzips it to a tempdir,
+removes any carriage returns found in /resources, and rezips it to the
+original filename.
+"""
+
+import zipfile
+import tempfile
+import os
+import sys
+import re
+
+
+def _write_files(outzip, contentsdir, verbose=False):
+    for dirpath, dirnames, filenames in os.walk(contentsdir):
+        for fn in filenames:
+            fn_abs = os.path.join(dirpath, fn)
+            zip_path = os.path.relpath(fn_abs, contentsdir)
+            outzip.write(fn_abs, zip_path)
+
+
+def _edit_files(indir, verbose=False):
+    to_edit = []
+    print(os.listdir(os.path.join(indir, 'opentrons')))
+    for root, _2, filenames in os.walk(os.path.join(indir,
+                                                    'opentrons',
+                                                    'resources')):
+        to_edit.extend([os.path.join(root, fn) for fn in filenames])
+    if verbose:
+        print("File edit list: {}".format(to_edit))
+    for fn in to_edit:
+        # We don’t need to edit python files
+        if verbose:
+            sys.stdout.write("Checking {}: ".format(fn))
+        if fn.endswith('.py') or fn.endswith('.pyc'):
+            if verbose:
+                print("Skipped (is python file)")
+            continue
+        # Or python files that don’t look like python files
+        try:
+            first = open(fn).readline()
+        except UnicodeDecodeError:  # (or binary files)
+            print("Skipped (is most likely binary)")
+            continue
+        else:
+            if first.startswith('#!') and 'python' in first:
+                if verbose:
+                    print("Skipped (is shebanged python file)")
+                continue
+        contents = open(fn, 'rb').read()
+        new = re.sub(b'\r\n', b'\n', contents)
+        if verbose:
+            if len(contents)-len(new) > 0:
+                print("Edited (found {} <cr>)"
+                      .format(len(contents)-len(new)))
+            else:
+                print("Nothing to do")
+        open(fn, 'wb').write(new)
+
+
+def rewrite_egg(infile, outfile, verbose=False):
+    """ Rewrite any <cr> in the resources/ subdirectory of the specified egg.
+
+    filelike should be an open filelike object holding the egg.
+    """
+    with tempfile.TemporaryDirectory() as tempdir:
+        with zipfile.ZipFile(infile, 'r') as inzip:
+            if verbose:
+                print("Extracting {} to {}".format(inzip, tempdir))
+            inzip.extractall(tempdir)
+        print("Editing files in {}".format(tempdir))
+        _edit_files(tempdir, verbose)
+        with zipfile.ZipFile(outfile, 'w') as outzip:
+            if verbose:
+                print("Zipping {} to {}".format(tempdir, outzip))
+            _write_files(outzip, tempdir, verbose)
+
+
+if __name__ == '__main__':
+    import argparse
+    parser = argparse.ArgumentParser(
+        description='Create a new opentrons egg without <cr> in shell scripts')
+    parser.add_argument(
+        'in_whl', metavar='INWHL', type=str,
+        help='The wheel file, or directory containing the egg file, to read')
+    parser.add_argument(
+        'out_whl', metavar='OUTWHL', type=str,
+        help='The wheel file to write. If not specified, execute in place',
+        nargs='?',
+        default=None)
+    parser.add_argument('-v', '--verbose', help='Print extra status messages',
+                        action='store_true')
+    args = parser.parse_args()
+    if os.path.isdir(args.in_whl):
+        files = os.listdir(args.in_whl)
+        for f in files:
+            if re.match('opentrons-.*.whl', f):
+                in_whl = os.path.join(args.in_whl, f)
+                break
+        else:
+            raise OSError("Couldn't find anything that looked like a wheel in {}"
+                          .format(args.in_whl))
+    else:
+        in_whl = args.in_whl
+    if None is args.out_whl:
+        out_whl = in_whl
+    else:
+        out_whl = args.out_whl
+
+    rewrite_egg(in_whl, out_whl, args.verbose)