run_tests
templatetags/__init__.py
tests/__init__.py
tests/django-headline.py
tests/font.ttf
tests/settings.py
@@ -1,294 +1,406 @@
-# -*- coding: utf-8 -*-
-#
-# Copyright (C) 2009 Igor "SkAZi" Potapov <igor@poatpoff.org>
-# GNU General Public License
-#
-# http://github.com/SkAZi/django-headline/
-
-from django import template
-from django.conf import settings
-from django.utils.encoding import smart_str
-from django.utils.safestring import mark_safe
-from django.template import TemplateSyntaxError
-
-import Image, ImageFont, ImageDraw, ImageChops
-from os import path
-try:
- from hashlib import md5
-except:
- from md5 import new as md5
-import re
-
-HEADLINE_CACHE_DIR = getattr(settings, 'HEADLINE_CACHE_DIR', 'upload')
-HEADLINE_NO_CACHE = getattr(settings, 'HEADLINE_NO_CACHE', False)
-HEADLINE_FONTS_DIR = getattr(settings, 'HEADLINE_FONTS_DIR', 'fonts')
-HEADLINE_CLASSES = getattr(settings, 'HEADLINE_CLASSES', {})
-HEADLINE_TEMPLATE = getattr(settings, 'HEADLINE_TEMPLATE',
- u"""<img alt="%(text)s" src="%(url)s" class="png" width="%(width)s" height="%(height)s" />""")
-HEADLINE_PNG_OPTIMIZER = getattr(settings, 'HEADLINE_PNG_OPTIMIZER', False)
-
-AVIABLE_DECORATIONS = ('underline', 'strikeout')
-AVIABLE_SPLITTERS = ('br', 'all', 'none')
-
-ENTITIES = (
- (u'«', u'«'), (u'»', u'»'),
- (u'„', u'„'), (u'“', u'“'),
- (u'&lquo;', u'”'), (u'&ndash', u'–'),
- (u'—', u'—'), (u'&', u'&'),
- (u'"', u"\""), (u''', u"'"),
- (u'®', u'®'), (u'©', u'©'),
- (u'™', u'™'), (u'§', u'§'),
- (u'€', u'€'), (u' ', u' '),
- (u'’', u'’'), (u'″', u'″'),
- (u'≤', u'≤'), (u'≥', u'≥'),
- (u'<', u'<'), (u'>', u'>'),
- (u'\n', u''), (u'\t', u' '),
-)
-
-
-def _clean_text(text):
- """
- Clean text from extra spaces and replaces
- html-entities by unicode symbols
- """
- text = text.strip()
- for (s, r) in ENTITIES:
- text = re.sub(s, r, text)
- return text
-
-
-
-def _img_from_text(text, font, size=12, color='#000', decoration=[]):
- """
- Draws text with font, size, color and decoration parameters.
- Caches images and returns (html or object, width, size) of
- new or exists image
- """
- image_path = path.join(settings.MEDIA_ROOT, HEADLINE_CACHE_DIR)
- font_path = path.join(settings.MEDIA_ROOT, HEADLINE_FONTS_DIR)
-
- id = "headline-%s" % md5(smart_str(''.join((text, font, size, color, ''.join(decoration))))).hexdigest()
- image_file = path.join(image_path, "%s.png" % id)
-
- size = int(size)
-
- if not path.isfile(image_file) or HEADLINE_NO_CACHE:
-
- font = ImageFont.truetype(path.join(font_path, font), size)
- width, height = font.getsize(text)
-
- ### Init surfaces
- image = Image.new("RGB", (width, height), (0,0,0))
- alpha = Image.new("L", image.size, "black")
- imtext = Image.new("L", image.size, 0)
- draw = ImageDraw.Draw(imtext)
-
- ### Real Drawings on alpha with white color
- draw.text((0, 0), text, font=font, fill="white")
- if 'underline' in decoration:
- draw.line((0 + size/20, height*3/4, width - size/20, height*3/4),
- fill="white", width=size/20)
- if 'strikeout' in decoration:
- draw.line((0 + size/20, height/2, width - size/20, height/2),
- fill="white", width=size/20)
-
- ### Alpha color black-magic
- alpha = ImageChops.lighter(alpha, imtext)
- solidcolor = Image.new("RGBA", image.size, color)
- immask = Image.eval(imtext, lambda p: 255 * (int(p != 0)))
- image = Image.composite(solidcolor, image, immask)
- image.putalpha(alpha)
- image.save(image_file, "PNG")
-
- if HEADLINE_PNG_OPTIMIZER:
- from os import system
- system(HEADLINE_PNG_OPTIMIZER % {"file": image_file})
-
- else:
- width, height = Image.open(image_file).size
-
- return "%s%s/%s.png" % \
- (settings.MEDIA_URL, HEADLINE_CACHE_DIR, id), width, height
-
-
-
-def _image_list(output, klass, splitter, object=False):
- """
- Pictures and html or objects generator by split options
- """
- if splitter:
- chunks = splitter.split(output)
- else:
- chunks = (output, )
-
- for text in chunks:
- text = _clean_text(text)
- if not text: continue
-
- url, width, height = _img_from_text(text, **klass)
- obj = { 'text': text,
- 'url': url,
- 'width': width,
- 'height': height }
-
- if object:
- yield obj
- else:
- yield HEADLINE_TEMPLATE % obj
-
-
-def _create_splitter(splitting):
- """
- Creates regular expression text splitter
- """
- if splitting == 'br':
- return re.compile(r"<br */?>"), "<br />"
- elif splitting == 'all':
- return re.compile(r"<br */?>| *"), " "
- else:
- return False, ""
-
-
-
-def _get_class( klass ):
- """
- Get or create real class parameters by options string
- """
- params = re.split(r", *", klass.strip('"'))
- typ = False
-
- if len(params) > 2:
- decoration = []
- for i in AVIABLE_DECORATIONS:
- if i in params:
- decoration.append(i)
- params.remove(i)
-
- klass = {
- 'font': params[0],
- 'size': params[1],
- 'color': params[2],
- 'decoration': decoration
- }
-
- if len(params) > 3:
- typ = params[3]
-
- return klass, typ
-
- elif params > 0:
- klass = params[0]
- if len(params) > 1:
- typ = params[1]
-
- try:
- return HEADLINE_CLASSES[klass], typ
- except:
- raise TemplateSyntaxError('There is no class: %s' % klass)
-
- return False, False
-
-
-
-register = template.Library()
-
-
-@register.filter(name="headline")
-def do_text_image_filter(value, klass):
- klass, typ = _get_class(klass)
- if not klass:
- raise TemplateSyntaxError('Headline filter needs in parameter' )
-
- return mark_safe(TextImageNode(value, klass, typ).show())
-
-do_text_image_filter.is_safe = True
-
-
-@register.tag(name="headline")
-def do_text_image_tag(parser, token):
-
- params = token.split_contents()
-
- klass, typ = _get_class(params[1])
-
- if not klass:
- raise TemplateSyntaxError('Headline tag needs in parameter' )
-
- if not typ:
- typ = "br"
-
- nodelist = parser.parse(('endheadline',))
- parser.delete_first_token()
-
- return TextImageNode(nodelist, klass, typ)
-
-
-class TextImageNode(template.Node):
-
- def __init__(self, nodelist, klass, splitting="br"):
- self.klass = klass
- self.splitter, self.joiner = _create_splitter(splitting)
- self.nodelist = nodelist
-
- def show(self):
- """ If filter """
- output = self.nodelist
- return self.joiner.join(
- _image_list( output, self.klass, self.splitter )
- )
-
- def render(self, context):
- """ If tag """
- output = self.nodelist.render(context)
- return self.joiner.join(
- _image_list( output, self.klass, self.splitter )
- )
-
-
-
-@register.tag(name="headlines")
-def do_text_images_tag(parser, token):
- params = token.split_contents()
-
- if len(params) > 4 and params[-3] == 'as':
- klass, typ = _get_class(params[-1])
-
- output = []
- for st in params[1:-3]:
- if st.startswith('"') and st.endswith('"'):
- output.append(st.strip('"'))
- else:
- output.append(parser.compile_filter(st))
-
- return TextImagesNode(output, params[-2], klass, typ)
-
- else:
- raise TemplateSyntaxError('Hedlines tag must be like {% headlines <list of vars> as <context_var> <class_or_params> %}' )
-
-
-class TextImagesNode(template.Node):
- def __init__(self, output, to, klass, splitting):
- self.klass = klass
- self.output = output
- self.to = to
- self.splitter, self.joiner = _create_splitter(splitting)
-
- def _flatten_data(self, output, context):
- """
- Generator of images from all-typed parameters
- """
- for token in output:
- if isinstance(token, (unicode, str)):
- token = (token, )
- else:
- token = token.resolve(context)
- if not isinstance(token, (list, tuple, dict)):
- token = (token, )
-
- for tk in token:
- for obj in _image_list( tk, self.klass, self.splitter, True):
- yield obj
-
- def render(self, context):
- context[self.to] = [item for item in self._flatten_data(self.output, context)]
- return ''
\ No newline at end of file
+# -*- coding: utf-8 -*-
+#
+# Copyright (C) 2009 Igor "SkAZi" Potapov <igor@poatpoff.org>
+# GNU General Public License
+#
+# http://github.com/SkAZi/django-headline/
+
+from django import template
+from django.conf import settings
+from django.utils.encoding import smart_str
+from django.utils.safestring import mark_safe
+from django.template import TemplateSyntaxError
+
+from htmlentitydefs import name2codepoint
+import Image, ImageFont, ImageDraw, ImageChops
+from os import path
+try:
+ from hashlib import md5
+except:
+ from md5 import new as md5
+import re
+
+HEADLINE_CACHE_DIR = getattr(settings, 'HEADLINE_CACHE_DIR', 'upload')
+HEADLINE_NO_CACHE = getattr(settings, 'HEADLINE_NO_CACHE', False)
+HEADLINE_FONTS_DIR = getattr(settings, 'HEADLINE_FONTS_DIR', 'fonts')
+HEADLINE_CLASSES = getattr(settings, 'HEADLINE_CLASSES', {})
+HEADLINE_TEMPLATE = getattr(settings, 'HEADLINE_TEMPLATE',
+ u"""<img alt="%(text)s" src="%(url)s" class="png" width="%(width)s" height="%(height)s" />""")
+HEADLINE_PNG_OPTIMIZER = getattr(settings, 'HEADLINE_PNG_OPTIMIZER', False)
+
+AVIABLE_SPLITTERS = ('br', 'all', 'none')
+AVIABLE_DECORATIONS = {
+ 'underline': 0,
+ 'strikeout': 0,
+ 'opacity': 1,
+ 'rotate': 0,
+}
+
+ENTITY_CONVERTER = re.compile(r'(?:&#x([a-fA-F\d]{1,4});)|(?:&#(\d{1,5});)|(?:&([a-zA-Z\d]+);)')
+
+
+
+def _convertentity(m):
+ """
+ Convert single entity into unicode character
+ """
+ ### If it is hex entity like ÿ
+ if m.group(1):
+ return unichr(int(m.group(1), 16))
+ ### If it is dec entity like ÿ
+ elif m.group(2):
+ return unichr(int(m.group(2)))
+ ### If it is named entity like «
+ elif m.group(3) and m.group(3) in name2codepoint:
+ return unichr(name2codepoint[m.group(3)])
+ ### Dont touch all other
+ else:
+ return m.group(0)
+
+
+
+def _clean_text(text):
+ """
+ Clean text from extra spaces and replaces
+ html-entities by unicode symbols
+ """
+ text = text.strip().replace(u'\n', u'').replace(u'\r', u'')
+ return ENTITY_CONVERTER.sub(_convertentity, text)
+
+
+
+
+def _img_from_text(text, font, size=12, color='#000', decoration={}):
+ """
+ Draws text with font, size, color and decoration parameters.
+ Caches images and returns (html or object, width, size) of
+ new or exists image
+ """
+
+ image_path = path.join(settings.MEDIA_ROOT, HEADLINE_CACHE_DIR)
+ font_path = path.join(settings.MEDIA_ROOT, HEADLINE_FONTS_DIR)
+
+ id = "headline-%s" % md5(smart_str(''.join((text, font, size.__str__(), color, decoration.__str__())))).hexdigest()
+ image_file = path.join(image_path, "%s.png" % id)
+
+
+ if not path.isfile(image_file) or HEADLINE_NO_CACHE:
+
+ size = int(size)
+ font = ImageFont.truetype(path.join(font_path, font), size)
+ width, height = font.getsize(text)
+
+ ### Init surfaces
+ image = Image.new("RGB", (width, height), (0,0,0))
+ alpha = Image.new("L", image.size, "black")
+ imtext = Image.new("L", image.size, 0)
+ draw = ImageDraw.Draw(imtext)
+
+ ### Real Drawings on alpha with white color
+ if decoration.has_key('opacity'):
+ opacity = float(decoration['opacity']) * 255
+ else:
+ opacity = 255
+
+ ### Draws text
+ draw.text((0, 0), text, font=font, fill=opacity)
+
+ ### Draws an underline
+ if decoration.has_key('underline'):
+ val = int(decoration['underline'])
+ draw.line((0 + size/20, height * 4 / 5 + val,
+ width - size/20, height * 4 / 5 + val),
+ fill=opacity, width=size / 20)
+
+ ### Draws an strikeout line
+ if decoration.has_key('strikeout'):
+ val = int(decoration['strikeout'])
+ draw.line((0 + size/20, height / 2 + val,
+ width - size/20, height / 2 + val),
+ fill=opacity, width=size / 20)
+
+ ### Alpha color black-magic
+ alpha = ImageChops.lighter(alpha, imtext)
+ solidcolor = Image.new("RGBA", image.size, color)
+ immask = Image.eval(imtext, lambda p: 255 * (int(p != 0)))
+ image = Image.composite(solidcolor, image, immask)
+ image.putalpha(alpha)
+
+ ### Rotation
+ if decoration.has_key('rotate') and decoration['rotate']:
+ angle = float(decoration['rotate'])
+ if angle == 90:
+ image = image.transpose(Image.ROTATE_90)
+ elif angle == 180:
+ image = image.transpose(Image.ROTATE_180)
+ elif angle == 270:
+ image = image.transpose(Image.ROTATE_270)
+ else:
+ # XXX: Bad rotation
+ # Really bicubic transformation works only
+ # when canvas doesn`t resize: last param is False
+ image = image.rotate(angle, Image.BICUBIC, True)
+ width, height = image.size
+
+ ### Save image
+ image.save(image_file, "PNG")
+
+ ### Optimize png with external tool
+ if HEADLINE_PNG_OPTIMIZER:
+ from os import system
+ system(HEADLINE_PNG_OPTIMIZER % {"file": image_file})
+
+ else:
+ ### We need just dimentions
+ width, height = Image.open(image_file).size
+
+ return "%s%s/%s.png" % \
+ (settings.MEDIA_URL, HEADLINE_CACHE_DIR, id), width, height
+
+
+
+def _image_list(output, klass, splitter, object=False):
+ """
+ Pictures and html or objects generator by split options
+ """
+ if splitter:
+ chunks = splitter.split(output)
+ else:
+ chunks = (output, )
+
+ for text in chunks:
+ text = _clean_text(text)
+ if not text: continue
+
+ url, width, height = _img_from_text(text, **klass)
+ obj = { 'text': text,
+ 'url': url,
+ 'width': width,
+ 'height': height }
+
+ if object:
+ yield obj
+ else:
+ yield HEADLINE_TEMPLATE % obj
+
+
+
+def _create_splitter(splitting):
+ """
+ Creates regular expression text splitter
+ """
+ if splitting == 'br':
+ return re.compile(r"<br */?>"), "<br />"
+ elif splitting == 'all':
+ return re.compile(r"<br */?>| *"), " "
+ else:
+ return False, ""
+
+
+
+def _get_class( klass ):
+ """
+ Get or create real class parameters by options string
+ """
+ params = re.split(r", *", klass.strip('"'))
+ cleaned = []
+ klass = False
+ typ = False
+
+ ### If it is render parameters string
+ if len(params) > 2:
+ decoration = {}
+
+ ### Remove all decorators, they can be anywhere
+ ### and add they into separate dict
+ for param in params:
+ param_unpack = param.split(':')
+ if param_unpack[0] in AVIABLE_DECORATIONS:
+ if len(param_unpack) > 1:
+ decoration[param_unpack[0]] = param_unpack[1]
+ else:
+ decoration[param_unpack[0]] = AVIABLE_DECORATIONS[param_unpack[0]]
+ else:
+ cleaned.append(param)
+
+ if(len(cleaned) < 3):
+ raise TemplateSyntaxError('Headline render parameters string must have at least 3 parameters (excepting decorations): %s' % ",".join(cleaned))
+ ### There we can have type of splitting too
+ elif len(cleaned) > 3:
+ typ = cleaned[3]
+
+ klass = {
+ 'font': cleaned[0],
+ 'size': int(cleaned[1]),
+ 'color': cleaned[2],
+ 'decoration': decoration
+ }
+
+ ### If it is class
+ elif params > 0:
+ try:
+ klass = HEADLINE_CLASSES[params[0]]
+ except:
+ raise TemplateSyntaxError('There is no class %s in HEADLINE_CLASSES definition.' % klass)
+
+ ### Back compatibility with v0.2
+ if isinstance(klass.setdefault('decoration', {}), (list, tuple)):
+ decoration = {}
+ for i in klass['decoration']:
+ decoration[i] = AVIABLE_DECORATIONS[i]
+ klass['decoration'] = decoration
+
+ ### Type of splitting can be here too
+ if len(params) > 1:
+ typ = params[1]
+
+ return klass, typ
+
+
+
+register = template.Library()
+
+
+
+@register.filter(name="headline")
+def do_text_image_filter(value, klass):
+ '''
+ Django filter, that generates png image with text
+ from variable and returns it`s html representation.
+
+ Examples:
+ {{ foo|headline:"font.ttf,20,#000" }}
+ {{ foo|headline:"font.ttf,20,#000,underline,all" }}
+ {{ foo|headline:"base,br" }}
+ '''
+ klass, typ = _get_class(klass)
+ if not klass:
+ raise TemplateSyntaxError('Headline filter needs in single render parameters string' )
+
+ return mark_safe(TextImageNode(value, klass, typ).render_text())
+
+do_text_image_filter.is_safe = True
+
+
+
+@register.tag(name="headline")
+def do_text_image_tag(parser, token):
+ '''
+ Django tag, that generates png image with text
+ from tag inside and returns it`s html representation.
+
+ Examples:
+ {% headline "font.ttf,20,opacity:0.5,#000" %}Big {{ foo }}{% endheadline %}
+ {% headline "font.ttf,20,#000,strikeout:-5,none" %}Big {{ foo }}{% endheadline %}
+ {% headline "base" %}Big {{ foo }}{% endheadline %}
+ '''
+ params = token.split_contents()
+
+ klass, typ = _get_class(params[1])
+
+ if not klass:
+ raise TemplateSyntaxError('Headline tag needs in single render parameters string' )
+
+ if not typ:
+ typ = "br"
+
+ nodelist = parser.parse(('endheadline',))
+ parser.delete_first_token()
+
+ return TextImageNode(nodelist, klass, typ)
+
+
+
+@register.tag(name="headlines")
+def do_text_images_tag(parser, token):
+ '''
+ Django tag, that generates png images with text
+ from tag parameters and returns it`s dict representations
+ into context variable.
+
+ Examples:
+ {% headlines foo_list bar_dict baz_var "And some text" as headers "font.ttf,20,#000" %}
+ {% headlines foo_list bar_dict baz_var "And some text" as headers "font.ttf,20,#000,all" %}
+ {% headlines foo_list bar_dict baz_var "And some text" as headers "base" %}
+ '''
+ params = token.split_contents()
+
+ if len(params) > 4 and params[-3] == 'as':
+ klass, typ = _get_class(params[-1])
+
+ output = []
+
+ for st in params[1:-3]:
+ ### If it is just a string
+ if st.startswith('"') and st.endswith('"'):
+ output.append(st.strip('"'))
+
+ ### Else it is context variable
+ else:
+ output.append(parser.compile_filter(st))
+
+ return TextImagesNode(output, params[-2], klass, typ)
+
+ else:
+ raise TemplateSyntaxError('Hedlines tag must have at least 4 parameters: one or more variables to render, "as" keyword, context variable and render parameters string. Like this: {% headlines <list of vars> as <context_var> <class_or_params> %}' )
+
+
+
+
+class TextImageNode(template.Node):
+
+ def __init__(self, nodelist, klass, splitting="br"):
+ self.klass = klass
+ self.splitter, self.joiner = _create_splitter(splitting)
+ self.nodelist = nodelist
+
+ def render_text(self):
+ """ If filter """
+ output = self.nodelist
+ return self.joiner.join(
+ _image_list( output, self.klass, self.splitter )
+ )
+
+ def render(self, context):
+ """ If tag """
+ output = self.nodelist.render(context)
+ return self.joiner.join(
+ _image_list( output, self.klass, self.splitter )
+ )
+
+
+
+class TextImagesNode(template.Node):
+ def __init__(self, output, to, klass, splitting):
+ self.klass = klass
+ self.output = output
+ self.to = to
+ self.splitter, self.joiner = _create_splitter(splitting)
+
+ def _flatten_data(self, output, context):
+ """
+ Generator of images from all-typed parameters
+ """
+ ### We walk thu all parameters
+ for token in output:
+
+ ### Gets just text strings
+ if isinstance(token, (unicode, str)):
+ token = (token, )
+
+ else:
+ ### And context variables
+ token = token.resolve(context)
+
+ ### If it single variable we transforms is into list
+ if not isinstance(token, (list, tuple, dict)):
+ token = (token, )
+
+ ### And then ruturns all of them in single stream
+ for tk in token:
+ for obj in _image_list( tk, self.klass, self.splitter, True):
+ yield obj
+
+ def render(self, context):
+ context[self.to] = [item for item in self._flatten_data(self.output, context)]
+ return ''
templatetags/headline.py
TODO
24bd8daa6dc91bcfc5918750b5f61e1131cc9cde
88adb109ea53e21d93a5af9dc8e57205890a2926
Igor SkAZi Potapov
igor@plan-b.ru
http://github.com/SkAZi/django-headline/commit/8599e7dd80dd4c5678314550a97053842a8e5612
8599e7dd80dd4c5678314550a97053842a8e5612
2009-07-14T23:06:44-07:00
2009-07-14T23:06:44-07:00
Merge branch 'unstable'
2d73ef5ecf3686e96e3b88a5c001502001a08802
Igor SkAZi Potapov
igor@plan-b.ru