huTools.http API and function-based pyJasper Interface. Dependencie R…

…eduction

This patch introduces a HTTP-API menat to allow easy use on pure
Python and AppEngine plattforms. All other huTools modules should
reley on them instead of httplib2 and others.

pyjasper now comes with an non Object Oriented interface and also strips
timestamps from generated PDFs to ensure the same input generates EXACTLY
the same output.

use hujson instead of simplejson.

.gitignore

@@ -5,5 +5,5 @@
 dist/
 docs/_build/doctrees/*
 huTools.egg-info/
-testenv/
+pythonenv/
 build/

Makefile

@@ -1,5 +1,5 @@
 # setting the PATH seems only to work in GNUmake not in BSDmake
-PATH := ./testenv/bin:$(PATH)
+PATH := ./pythonenv/bin:$(PATH)
 
 default: dependencies check test examples
 
@@ -16,18 +16,20 @@ check:
 	-/usr/local/hudorakit/bin/hd_pylint huTools
 
 test:
-	PYTHONPATH=. python huTools/NetStringIO.py
-	PYTHONPATH=. python huTools/calendar/formats.py
-	PYTHONPATH=. python huTools/calendar/workdays.py
-	PYTHONPATH=. python huTools/checksumming.py
-	PYTHONPATH=. python huTools/humessaging.py
-	PYTHONPATH=. python huTools/luids.py
-	PYTHONPATH=. python huTools/obfuscation.py
-	PYTHONPATH=. python huTools/unicode.py
+	PYTHONPATH=. ./pythonenv/bin/python huTools/http/test.py
+	PYTHONPATH=. ./pythonenv/bin/python huTools/NetStringIO.py
+	PYTHONPATH=. ./pythonenv/bin/python huTools/calendar/formats.py
+	PYTHONPATH=. ./pythonenv/bin/python huTools/calendar/workdays.py
+	PYTHONPATH=. ./pythonenv/bin/python huTools/checksumming.py
+	PYTHONPATH=. ./pythonenv/bin/python huTools/humessaging.py
+	PYTHONPATH=. ./pythonenv/bin/python huTools/luids.py
+	PYTHONPATH=. ./pythonenv/bin/python huTools/obfuscation.py
+	PYTHONPATH=. ./pythonenv/bin/python huTools/unicode.py
+	PYJASPER_SERVLET_URL=http://127.0.0.1:8000/pyJasper/jasper.py PYTHONPATH=. ./pythonenv/bin/python huTools/pyjasper.py
 
 coverage: dependencies
 	printf '.*/tests/.*\n.*test.py\n' > .figleaf-exclude.txt
-	printf '/usr/local/lib/.*\n/opt/.*\ntestenv/.*\n' >> .figleaf-exclude.txt
+	printf '/usr/local/lib/.*\n/opt/.*\npythonenv/.*\n' >> .figleaf-exclude.txt
 	printf '.*manage.py\n.*settings.py\n.*setup.py\n.*urls.py\n' >> .figleaf-exclude.txt
 	PYTHONPATH=. /usr/local/hudorakit/bin/hd_figleaf --ignore-pylibs huTools/NetStringIO.py
 	PYTHONPATH=. /usr/local/hudorakit/bin/hd_figleaf --ignore-pylibs huTools/ReReadingConfigParser.py
@@ -53,8 +55,8 @@ build: examples
 	python setup.py build
 
 dependencies:
-	virtualenv testenv
-	pip -q install -E testenv -r requirements.txt
+	virtualenv pythonenv
+	pip -q install -E pythonenv -r requirements.txt
 
 statistics:
 	sloccount --wide --details huTools | tee sloccount.sc
@@ -77,7 +79,7 @@ examples: doc/standards/examples/warenzugang.xml doc/standards/examples/kommiauf
 
 
 clean:
-	rm -Rf testenv build dist html test.db pylint.out sloccount.sc pip-log.txt
+	rm -Rf testenv pythonenv build dist html test.db pylint.out sloccount.sc pip-log.txt
 	find . -name '*.pyc' -delete
 
-.PHONY: build clean install upload check doc
+.PHONY: build clean install upload check doc docs test

doc/development.markdown

@@ -103,7 +103,7 @@ Use global unique identifiers where ever possible. `huTools.luids.guid128()` cre
 * Always test Iñtërnâtiônàlizætiøn by putting strange strings into input fields
 * Always test `<script>alert("XSS");</script> & <bold>Co</bold>` by putting strange strings into input fields
 * use [huTools](http://hudora.github.com/huTools/) where appropriate
-
+* `Iñtërnâtiônàlizætiøn <script>alert("XSS");</script> %+'"<!--` might be a goot test string.
 
 ## Django Specifica
 

doc/standards/examples/wms2logos_kommiauftrag.xml

@@ -10,7 +10,7 @@
       <Liefertermin>20091125</Liefertermin>
       <Terminart>BIS</Terminart>
       <Textcode1>8</Textcode1>
-      <Auftragstext>abholer: Abholung durch Spedition Dornseif$hebebuehne: True$</Auftragstext>
+      <Auftragstext>abholer: Abholung durch Spedition Dornseif$avisieren_unter: +49123456$hebebuehne: True$</Auftragstext>
       <Textcode2>2</Textcode2>
       <Kommissioniertext>etiketten_speicherort: c:\desktop\Programm Data\schrecklicher windows pfad$sortenrein: True$packhoehe: 1100</Kommissioniertext>
       <EmpfaengerILN/>

huTools/http/__init__.py

@@ -0,0 +1,85 @@
+#!/usr/bin/env python
+# encoding: utf-8
+"""
+huTools/http/__init__.py
+
+Provides a simple interface for doing HTTP requests on stock Python and on Google Appengine. Provides
+unicode aware application/x-www-form-urlencoded encoding and an do multipart/form-data (file upload) 
+without aditional code. On Google Appengine it incerass the timeout from 5 seconds to 10.
+
+Usage is without suprises::
+
+    >>> status, header, body = fetch('http://www.postbin.org/o0ds54',
+                                     {'küh': 'Iñtërnâtiônàlizætiøn', 'just a test': 212},
+                                     headers={'X-Foo': 'Bar'})
+
+File Upload just works::
+
+    >>> status, header, body = fetch('http://www.postbin.org/o0ds54',
+                                     {'hosts': open('/etc/hosts', 'r')}, 'POST')
+
+"""
+# Created by Maximillian Dornseif on 2010-10-24.
+# Copyright (c) 2010 HUDORA. All rights reserved.
+
+from huTools.http import tools
+import logging
+import poster_encode
+import urllib
+import urlparse
+
+try:
+    from engine_appengine import request
+except ImportError:
+    from engine_httplib2 import request
+
+
+def fetch(url, content='', method='GET', credentials=None, headers=None, multipart=False):
+    """Does a HTTP request with method `method` to `url`. 
+    
+    Returns (status, headers, content) whereas `status` is an integer status code, `headers` is a dict
+    containing the headers sent by the server and `content` is the body of the http response.
+    
+    Parameters to fetch are::
+    
+    * `url` is the fully qualified request URL. It may contain query prameters.
+    * `content` is the request body to be sent. It may be a dict which for all requests expect POST
+      is converted to query parameters. If there are query parameters already in the `url` they are merged
+      with `content`. For POST requests the data is encoded as application/x-www-form-urlencoded
+      or multipart/form-data and encoded. If the parameter `multipart` is `True` or if one of the values
+      in content has a `name` attribute (which is the case for file objects) multipart encoding is choosen.
+    * `headers` is a dict of header values
+    * `credentials` is reserved for future use
+    """
+
+    if not headers:
+        headers = {}
+    if method == 'POST':
+        if hasattr(content, 'items'):
+            # we assume content is a dict which needs to be encoded
+            # decide to use multipart/form-data encoding or application/x-www-form-urlencoded
+            for v in content.values():
+                if hasattr(v, 'name'):
+                    multipart = True
+            if multipart:
+                datagen, mp_headers = poster_encode.multipart_encode(content)
+                headers.update(mp_headers)
+                content = "".join(datagen)
+            else:
+                headers.update({'content-type': 'application/x-www-form-urlencoded'})
+                content = tools.urlencode(content)
+    else:
+        # url parmater encoding
+        if hasattr(content, 'items'):
+            scheme, netloc, path, params, query, fragment = urlparse.urlparse(url)
+            qdict = urlparse.parse_qs(query)
+            # ugly Unicode issues, see http://bugs.python.org/issue1712522
+            qdict.update(content)
+            query = tools.urlencode(qdict)
+            url = urlparse.urlunparse((scheme, netloc, path, params, query, fragment))
+            content = ''
+    # convert all header values to strings (what about unicode?)
+    for k, v in headers.items():
+        headers[k] = str(v)
+
+    return request(url, method, content, headers)