diff --git a/generator/file_parser.py b/generator/file_parser.py index a0948f7..5156219 100644 --- a/generator/file_parser.py +++ b/generator/file_parser.py @@ -13,19 +13,40 @@ class WebPage(object): self.name=os.path.split(path)[-1] self.variables={} self.list_of_lang=set() - self.reserved=["model"] - self.init_reserved() + self.reserved=set(["model"]) + self.reserved_prefix="reserved_" + self.content=[] - def init_reserved(self): - for key in reserved: - self.setattr(key,"") + def init_reserved(self,*others): + "Init all the reserved variables, note that no protections are used, so use them at your own risk" + for i in others: + self.reserved.add(i) + for key in self.reserved: + setattr(self,key,"") + + def set_reserved(self,var): + setattr(self,self.reserved_prefix+key,"") + + def get_reserved(self,var): + getattr(self,var) + + def add_content(self,text,lang): + self.content.append([text,lang]) + + def get_next_line(self,lang="*"): + "Get a line of text, with a filter if needed" + pass + + def get_lang(self,lang): + "Will make sense if I ever use a translation table and not a clusterf*ck of strings" + return lang def add_variable(self,var,value): """Add the folowing variable and update the necessary constants Takes a variable (with an eventual langague tag) and it's value""" var,lang,value=create_lang(var,value) - if var in reserved: - self.setattr(var,value) + if var in self.reserved: + setattr(self,var,value) else: if var in self.variables: self.variables[var][lang]=value @@ -45,29 +66,56 @@ def create_lang(var,value): def parse_file(file_name): "Parse a file and return a webpage object" page=WebPage(file_name) - isconfig=True # The firsts lines are config lines + page.init_reserved() with open(file_name) as f: - while True: - line=f.readline() - if not line: - break - try: + try: + while True: # Parse the config file + line=f.readline() + if not line: # Wait for the last line + break if line.startswith("----"): # Not really a good practice, but meh - isconfig=False - if isconfig and line!=os.linesep: # Now parsing config files + break + if line!=os.linesep: # Now parsing config files var,value=re_config_line.match(line).groups() page.add_variable(var,value) - else: # The config lines are now parsed, will now enter the nightmare of standart lines + while True: # The config lines are now parsed, will now enter the nightmare of standart lines + line=f.readline() + if not line: # Wait for the last line + break + match=re_text_line.match(line) #Will always match since there is a .* in the regex + beg_lang,end_lang,text=match.groups() + used_langs={} + page_text=[] + if beg_lang: #Will now add a lang to witch the programe should write + if beg_lang in used_langs: + used_langs[beg_lang]+=1 + else: + used_langs[beg_lang]=1 + elif end_lang: + if end_lang in used_langs: + used_langs[end_lang]-=1 + else: + used_langs[end_lang]=0 # This should never happen, but...users + elif text: + line_langs=[] + for l in used_langs: + if used_langs[l]>0: + line_langs.append(l) + if len(used_langs)==0: # If no langs are used, print in every lang + used_langs=["*"] + page.add_content(text,line_langs) - except re.error: - print "Error parsing",file_name,"contain a non parsable line:" - print " >",line - except: - traceback.print_exc() + except re.error: + print "Error parsing",file_name,"contain a non parsable line:" + print " >",line + except: + traceback.print_exc() + return page re_config_line=re.compile("(?P.+): (?P.*)") -re_text_line=re.compile("") +re_text_line=re.compile("__(?P\w+)__|__/(?P\w+)__|(?P.*)") if __name__ == '__main__': - - parse_file("sites/example_website/_config.txt") + page=parse_file("sites/example_website/_config.txt") + print page.variables + print page.list_of_lang diff --git a/sites/example_website/_config.txt b/sites/example_website/_config.txt index 82028ce..1074a70 100644 --- a/sites/example_website/_config.txt +++ b/sites/example_website/_config.txt @@ -10,3 +10,7 @@ generation: dynamic Be wary that _ is a reserved character Technically those lines are not comments but since the text content of config files is ignored, it acts like one + +__en__ + +__/en__