"""
Define the HtmlPreprocessor class to generate a single HTML file from many source files.
"""
import base64
import re
import zlib
import json
from pathlib import Path
[docs]
class HtmlPreprocessor():
"""
Recursively substitute and insert source files to produce a single HTML file.
Source files contain text with directives in the form: <<hpp_directive arg [flag1] [flag2]>>
Parameters
----------
start_filename : str
The file to begin processing from.
output_filename : str
The path to the newly merged HTML file.
search_path : str[]
List of directory names to search for files.
allow_overwrite : bool
If true, overwrite the output file if it exists.
var_dict : dict
Dictionary of variable names and values referenced by hpp_pyvar directives.
json_dumps_default : function
Passed to json.dumps() as the "default" parameter that gets
called for objects that can't be serialized.
verbose : bool
If True, print some status messages to stdout.
Attributes
----------
_start_path : Path
A Path object to help with testing functions for the provided start_filename.
_start_filename : str
The path to the file to begin processing from, normally an HTML file.
_search_path : Path
List of Path objects to search for included files.
_output_filename : str
The path to the newly merged HTML file.
_var_dict : dict
Dictionary of variable names and values referenced by hpp_pyvar directives.
_allow_overwrite : bool
If true, overwrite the output file if it exists.
_verbose : bool
If True, print some status messages to stdout.
_loaded_filenames : list
List of filenames loaded so far, to determine if a file has been already loaded.
_json_dumps_default : function
Passed to json.dumps() as the "default" parameter that gets
called for objects that can't be serialized.
Notes
-----
Recognized Directives:
:code:`<<hpp_insert path/to/file [compress] [dup]>>`
Paste :code:`path/to/file` verbatim into the text.
:code:`<<hpp_script path/to/script [dup]>>`
Insert :code:`path/to/script` between script tags.
:code:`<<hpp_style path/to/css [dup]>>`
Insert :code:`path/to/css` between style tags.
:code:`<<hpp_bin2b64 path/to/file [dup]>>`
Convert a binary file to a b64 string and insert it.
:code:`<<hpp_pyvar variable_name [compress]>>`
Insert the string value of the named Python variable. If the referenced
variable is non-primitive, it's converted to JSON.
Flags:
:code:`compress` :
The replacement content will be compressed and converted to
a base64 string. It's up to the JavaScript code to decode and uncompress it.
:code:`dup` :
If a file has already been included once, it will be ignored on subsequent inclusions
unless the dup flag is used.
Commented directives (:code:`//`, :code:`/* */`, or :code:`<!-- -->`) will replace the
entire comment. When a directive is commented, it can only be on a single line or the
comment-ending characters will not be replaced.
All paths in the directives are relative to the directory that the start file
is located in (or if start_path was specified) unless it is absolute.
Nothing is written until every directive has been successfully processed.
"""
[docs]
def __init__(self, start_filename: str, output_filename: str, search_path=[],
allow_overwrite=False, var_dict: dict = None, json_dumps_default=None,
verbose=False):
"""
Configure the preprocessor and validate file paths.
"""
self._start_path = Path(start_filename)
if self._start_path.is_file() is False:
raise FileNotFoundError(f"Error: {self._start_path} not found")
self._search_path = []
for path_name in search_path:
self._search_path.append(Path(path_name))
# Put location of start file after manually-specified paths:
self._search_path.append(self._start_path.resolve().parent)
# Current folder:
self._search_path.append(Path('.'))
output_path = Path(output_filename)
if output_path.is_file() and not allow_overwrite:
raise FileExistsError(f"Error: {output_filename} already exists")
self._start_filename = start_filename
self._output_filename = output_filename
self._allow_overwrite = allow_overwrite
self._var_dict = var_dict
self._json_dumps_default = json_dumps_default
self._verbose = verbose
# Keep track of filenames already loaded, to make sure
# we don't unintentionally include the exact same file twice.
self._loaded_filenames = []
self.msg("HtmlProcessor object created.")
[docs]
def find_file(self, filename: str) -> str:
"""
Check specified locations for the provided filename.
Parameters
----------
filename : str
The path to the text file to locate.
Returns
-------
str
The full path to the existing file if found.
"""
file_path = Path(filename)
if file_path.is_absolute():
if file_path.is_file():
return filename
else:
for path in self._search_path:
test_path = Path(path / filename)
if test_path.is_file():
return test_path
raise FileNotFoundError(f"Error: {filename} not found")
[docs]
def load_file(self, filename: str, rlvl=0, binary=False, allow_dup=False) -> str:
"""
Open and read the specified text file.
Parameters
----------
filename : str
The path to the text file to read.
rlvl : int
Recursion level to help with indentation when verbose is enabled.
binary : bool
True if the file is to be opened in binary mode and converted to a base64 str.
allow_dup : bool
If False, return an empty string for a filename that's been previously loaded.
Returns
-------
str
The complete contents of the file.
"""
pathname = self.find_file(filename)
if pathname in self._loaded_filenames and not allow_dup:
self.msg(f"Ignoring previously-loaded file {filename}.", rlvl)
return ""
self._loaded_filenames.append(pathname)
self.msg(f"Loading file {pathname}.", rlvl)
if binary:
with open(pathname, 'rb') as f:
file_contents = str(base64.b64encode(f.read()).decode('UTF-8'))
else:
with open(pathname, 'r', encoding='UTF-8') as f:
file_contents = str(f.read())
return file_contents
[docs]
def msg(self, msg: str, rlvl=0):
"""
Print a message to stdout if self.verbose is True.
Parameters
----------
msg : str
The message to print.
rlvl : int
Recursion level to help with indentation when verbose is enabled.
"""
if self._verbose:
print(rlvl * '--' + msg)
[docs]
def parse_contents(self, contents: str, rlvl=0) -> str:
"""
Find the preprocessor directives in the file and replace them with the desired content.
Will recurse if directives are also found in the new content.
Parameters
----------
contents : str
The contents of a preloaded text file.
rlvl : int
Recursion level to help with indentation when verbose is enabled.
Returns
-------
str
The complete contents represented as a base64 string.
"""
# Find all possible keywords:
keyword_regex = \
r'(//|/\*|<\!--)?\s*<<\s*hpp_(\S+)\s+(\S+)(\s+compress|\s+dup)*\s*>>(\*/|-->)?'
matches = re.finditer(keyword_regex, contents)
rlvl += 1
new_content = None
for found_directive in matches:
full_match = found_directive.group(0)
# comment_start = found_directive.group(1)
keyword = found_directive.group(2)
arg = found_directive.group(3)
flags = {'compress': False, 'dup': False}
if found_directive.group(4) is not None:
flags['compress'] = True if 'compress' in found_directive.group(4) else False
flags['dup'] = True if 'dup' in found_directive.group(4) else False
do_compress = False # Change below with directives where it's allowed
self.msg(f"Handling {keyword} directive.", rlvl)
if keyword == 'insert':
# Recursively insert a plain text file which may also have hpp directives
new_content = self.parse_contents(self.load_file(arg, rlvl=rlvl,
allow_dup=flags['dup']), rlvl)
if new_content != "":
do_compress = True if flags['compress'] else False
elif keyword == 'script':
# Recursively insert a JavaScript file which may also have hpp directives
new_content = self.parse_contents(self.load_file(arg, rlvl=rlvl,
allow_dup=flags['dup']), rlvl)
if new_content != "":
new_content = f'<script type="text/javascript">\n{new_content}\n</script>'
elif keyword == 'style':
# Recursively insert a CSS file which may also have hpp directives
new_content = '<style type="text/css">\n' + \
self.parse_contents(self.load_file(arg, rlvl=rlvl,
allow_dup=flags['dup']), rlvl) + '\n</style>'
elif keyword == 'bin2b64':
new_content = self.load_file(arg, binary=True, allow_dup=flags['dup'], rlvl=rlvl)
elif keyword == 'pyvar':
if arg in self._var_dict:
val = self._var_dict[arg]
if type(val) in (str, bool, int, float):
# Use string representations of primitive types
new_content = str(self._var_dict[arg])
do_compress = True if flags['compress'] else False
else:
raw_data = json.dumps(val, default=self._json_dumps_default)
if flags['compress']:
self.msg("Compressing content.", rlvl)
compressed_content = zlib.compress(raw_data.encode('UTF-8'))
new_content = str(base64.b64encode(compressed_content).decode("UTF-8"))
else:
new_content = raw_data
else:
raise ValueError(f"Can't substitute for undefined variable {arg}")
else:
# Bad keyword
raise ValueError(f"Unknown HTML preprocessor directive hpp_{keyword}")
if do_compress:
self.msg("Compressing new content.", rlvl)
new_content = str(base64.b64encode(zlib.compress(new_content)).decode("UTF-8"))
if new_content is not None:
self.msg(f"Replacing directive '{full_match}' with new content.", rlvl)
# contents = re.sub(full_match, new_content, contents)
contents = contents.replace(full_match, new_content)
return contents
[docs]
def run(self) -> None:
"""
Initialize the preprocessor, then save the result as a new file.
"""
new_html_content = self.parse_contents(self.load_file(self._start_filename))
path = Path(self._output_filename)
if path.is_file() and not self._allow_overwrite:
raise FileExistsError(f"Error: {self._output_filename} already exists")
output_file = open(self._output_filename, "w", encoding='UTF-8')
output_file.write(new_html_content)
output_file.close()
