datamule 0.381__py3-none-any.whl → 1.0.0__py3-none-any.whl

datamule/document.py

@@ -0,0 +1,278 @@
+import json
+import csv
+from .helper import convert_to_dashed_accession
+import re
+from doc2dict import xml2dict, txt2dict
+from doc2dict.mapping import flatten_hierarchy
+from .mapping_dicts import txt_mapping_dicts
+from .mapping_dicts import xml_mapping_dicts
+from selectolax.parser import HTMLParser
+
+class Document:
+    def __init__(self, type, filename):
+        self.type = type
+        self.path = filename
+
+        self.data = None
+        self.content = None
+
+
+    def load_content(self,encoding='utf-8'):
+        with open(self.path, 'r',encoding=encoding) as f:
+            self.content = f.read()
+
+    def _load_text_content(self):
+        with open(self.path) as f:
+            return f.read().translate(str.maketrans({
+                '\xa0': ' ', '\u2003': ' ',
+                '\u2018': "'", '\u2019': "'",
+                '\u201c': '"', '\u201d': '"'
+            }))
+
+    # will deprecate this when we add html2dict
+    def _load_html_content(self):
+        with open(self.path,'rb') as f:
+            parser = HTMLParser(f.read(),detect_encoding=True,decode_errors='ignore')
+        
+        # Remove hidden elements first
+        hidden_nodes = parser.css('[style*="display: none"], [style*="display:none"], .hidden, .hide, .d-none')
+        for node in hidden_nodes:
+            node.decompose()
+        
+        blocks = {'p', 'div', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'article', 'section', 'li', 'td'}
+        lines = []
+        current_line = []
+        
+        def flush_line():
+            if current_line:
+                # Don't add spaces between adjacent spans
+                lines.append(''.join(current_line))
+                current_line.clear()
+        
+        for node in parser.root.traverse(include_text=True):
+            if node.tag in ('script', 'style', 'css'):
+                continue
+                
+            if node.tag in blocks:
+                flush_line()
+                lines.append('')
+                
+            if node.text_content:
+                text = node.text_content.strip()
+                if text:
+                    if node.tag in blocks:
+                        flush_line()
+                        lines.append(text)
+                        lines.append('')
+                    else:
+                        # Only add space if nodes aren't directly adjacent
+                        if current_line and not current_line[-1].endswith(' '):
+                            if node.prev and node.prev.text_content:
+                                if node.parent != node.prev.parent or node.prev.next != node:
+                                    current_line.append(' ')
+                        current_line.append(text)
+        
+        flush_line()
+        
+        text = '\n'.join(lines)
+        while '\n\n\n' in text:
+            text = text.replace('\n\n\n', '\n\n')
+        
+        return text.translate(str.maketrans({
+            '\xa0': ' ', '\u2003': ' ',
+            '\u2018': "'", '\u2019': "'",
+            '\u201c': '"', '\u201d': '"'
+        }))
+
+    def _load_file_content(self):
+        if self.path.suffix =='.txt':
+            self.content = self._load_text_content()
+        elif self.path.suffix in ['.html','.htm']:
+            self.content =  self._load_html_content()
+        else:
+            raise ValueError(f"Unsupported file type: {self.path.suffix}")
+
+
+    def contains_string(self, pattern):
+        """Currently only works for .htm, .html, and .txt files"""
+        if self.path.suffix in ['.htm', '.html', '.txt']:
+            if self.content is None:
+                self.content = self._load_file_content(self.path)
+            return bool(re.search(pattern, self.content))
+        return False
+
+    # Note: this method will be heavily modified in the future
+    def parse(self):
+        mapping_dict = None
+
+        if self.path.suffix == '.xml':
+            if self.type in ['3', '4', '5']:
+                mapping_dict = xml_mapping_dicts.dict_345
+
+            self.load_content()
+            self.data = xml2dict(content=self.content, mapping_dict=mapping_dict)
+        # will deprecate this when we add html2dict
+        elif self.path.suffix in ['.htm', '.html','.txt']:
+            self._load_file_content()
+
+            if self.type == '10-K':
+                mapping_dict = txt_mapping_dicts.dict_10k
+            elif self.type == '10-Q':
+                mapping_dict = txt_mapping_dicts.dict_10q
+            elif self.type == '8-K':
+                mapping_dict = txt_mapping_dicts.dict_8k
+            elif self.type == 'SC 13D':
+                mapping_dict = txt_mapping_dicts.dict_13d
+            elif self.type == 'SC 13G':
+                mapping_dict = txt_mapping_dicts.dict_13g
+            
+            self.data = txt2dict(content=self.content, mapping_dict=mapping_dict)
+        return self.data
+    
+    def write_json(self, output_filename=None):
+        if not self.data:
+            self.parse()
+            
+        if output_filename is None:
+            output_filename = f"{self.path.rsplit('.', 1)[0]}.json"
+            
+        with open(output_filename, 'w',encoding='utf-8') as f:
+            json.dump(self.data, f, indent=2)
+
+    def write_csv(self, output_filename=None, accession_number=None):
+        self.parse()
+
+        if output_filename is None:
+            output_filename = f"{self.path.rsplit('.', 1)[0]}.csv"
+
+        with open(output_filename, 'w', newline='') as csvfile:
+            if not self.data:
+                return output_filename
+
+            has_document = any('document' in item for item in self.data)
+            
+            if has_document and 'document' in self.data:
+                writer = csv.DictWriter(csvfile, ['section', 'text'], quoting=csv.QUOTE_ALL)
+                writer.writeheader()
+                flattened = self._flatten_dict(self.data['document'])
+                for section, text in flattened.items():
+                    writer.writerow({'section': section, 'text': text})
+            else:
+                fieldnames = list(self.data[0].keys())
+                if accession_number:
+                    fieldnames.append('Accession Number')
+                writer = csv.DictWriter(csvfile, fieldnames, quoting=csv.QUOTE_ALL)
+                writer.writeheader()
+                for row in self.data:
+                    if accession_number:
+                        row['Accession Number'] = convert_to_dashed_accession(accession_number)
+                    writer.writerow(row)
+
+        return output_filename
+    
+    def _document_to_section_text(self, document_data, parent_key=''):
+        items = []
+        
+        if isinstance(document_data, dict):
+            for key, value in document_data.items():
+                # Build the section name
+                section = f"{parent_key}_{key}" if parent_key else key
+                
+                # If the value is a dict, recurse
+                if isinstance(value, dict):
+                    items.extend(self._document_to_section_text(value, section))
+                # If it's a list, handle each item
+                elif isinstance(value, list):
+                    for i, item in enumerate(value):
+                        if isinstance(item, dict):
+                            items.extend(self._document_to_section_text(item, f"{section}_{i+1}"))
+                        else:
+                            items.append({
+                                'section': f"{section}_{i+1}",
+                                'text': str(item)
+                            })
+                # Base case - add the item
+                else:
+                    items.append({
+                        'section': section,
+                        'text': str(value)
+                    })
+        
+        return items
+
+    # we'll modify this for every dict
+    def _flatten_dict(self, d, parent_key=''):
+        items = {}
+        
+        if isinstance(d, list):
+            return [self._flatten_dict(item) for item in d]
+                
+        for k, v in d.items():
+            new_key = f"{parent_key}_{k}" if parent_key else k
+            
+            if isinstance(v, dict):
+                items.update(self._flatten_dict(v, new_key))
+            else:
+                items[new_key] = str(v)
+                    
+        return items
+   
+   # this will all have to be changed. default will be to flatten everything
+    def __iter__(self):
+        if not self.data:
+            self.parse()
+
+        # Let's remove XML iterable for now
+
+        # Handle text-based documents
+        if self.path.suffix in ['.txt', '.htm', '.html']:
+            document_data = self.data
+            if not document_data:
+                return iter([])
+                
+            # Find highest hierarchy level from mapping dict
+            highest_hierarchy = float('inf')
+            section_type = None
+            
+            if self.type in ['10-K', '10-Q']:
+                mapping_dict = txt_mapping_dicts.dict_10k if self.type == '10-K' else txt_mapping_dicts.dict_10q
+            elif self.type == '8-K':
+                mapping_dict = txt_mapping_dicts.dict_8k
+            elif self.type == 'SC 13D':
+                mapping_dict = txt_mapping_dicts.dict_13d
+            elif self.type == 'SC 13G':
+                mapping_dict = txt_mapping_dicts.dict_13g
+            else:
+                return iter([])
+                
+            # Find section type with highest hierarchy number
+            highest_hierarchy = -1  # Start at -1 to find highest
+            for mapping in mapping_dict['rules']['mappings']:
+                if mapping.get('hierarchy') is not None:
+                    if mapping['hierarchy'] > highest_hierarchy:
+                        highest_hierarchy = mapping['hierarchy']
+                        section_type = mapping['name']
+                        
+            if not section_type:
+                return iter([])
+                
+            # Extract sections of the identified type
+            def find_sections(data, target_type):
+                sections = []
+                if isinstance(data, dict):
+                    if data.get('type') == target_type:
+                        sections.append({
+                            'item': data.get('text', ''),
+                            'text': flatten_hierarchy(data.get('content', []))
+                        })
+                    for value in data.values():
+                        if isinstance(value, (dict, list)):
+                            sections.extend(find_sections(value, target_type))
+                elif isinstance(data, list):
+                    for item in data:
+                        sections.extend(find_sections(item, target_type))
+                return sections
+                
+            return iter(find_sections(document_data, section_type))
+            
+        return iter([])

datamule/downloader/downloader.py

@@ -0,0 +1,374 @@
+import asyncio
+import aiohttp
+import os
+from tqdm import tqdm
+from datetime import datetime
+from urllib.parse import urlencode
+import aiofiles
+import json
+import time
+from collections import deque
+
+from ..helper import identifier_to_cik, load_package_csv, fix_filing_url, headers
+from secsgml import parse_sgml_submission
+
+class RetryException(Exception):
+    def __init__(self, url, retry_after=601):
+        self.url = url
+        self.retry_after = retry_after
+
+class PreciseRateLimiter:
+    def __init__(self, rate, interval=1.0):
+        self.rate = rate  # requests per interval
+        self.interval = interval  # in seconds
+        self.token_time = self.interval / self.rate  # time per token
+        self.last_time = time.time()
+        self.lock = asyncio.Lock()
+
+    async def acquire(self):
+        async with self.lock:
+            now = time.time()
+            wait_time = self.last_time + self.token_time - now
+            if wait_time > 0:
+                await asyncio.sleep(wait_time)
+            self.last_time = time.time()
+            return True
+
+    async def __aenter__(self):
+        await self.acquire()
+        return self
+
+    async def __aexit__(self, exc_type, exc, tb):
+        pass
+
+class RateMonitor:
+    def __init__(self, window_size=1.0):
+        self.window_size = window_size
+        self.requests = deque()
+        self._lock = asyncio.Lock()
+    
+    async def add_request(self, size_bytes):
+        async with self._lock:
+            now = time.time()
+            self.requests.append((now, size_bytes))
+            while self.requests and self.requests[0][0] < now - self.window_size:
+                self.requests.popleft()
+    
+    def get_current_rates(self):
+        now = time.time()
+        while self.requests and self.requests[0][0] < now - self.window_size:
+            self.requests.popleft()
+        
+        if not self.requests:
+            return 0, 0
+        
+        request_count = len(self.requests)
+        byte_count = sum(size for _, size in self.requests)
+        
+        requests_per_second = request_count / self.window_size
+        mb_per_second = (byte_count / 1024 / 1024) / self.window_size
+        
+        return round(requests_per_second, 1), round(mb_per_second, 2)
+
+class Downloader:
+    def __init__(self):
+        self.headers = headers
+        self.limiter = PreciseRateLimiter(5)  # 10 requests per second
+        self.session = None
+        self.parse_filings = True
+        self.download_queue = asyncio.Queue()
+        self.rate_monitor = RateMonitor()
+        self.current_pbar = None
+        self.connection_semaphore = asyncio.Semaphore(5)
+    
+    def update_progress_description(self):
+        if self.current_pbar:
+            reqs_per_sec, mb_per_sec = self.rate_monitor.get_current_rates()
+            self.current_pbar.set_description(
+                f"Progress [Rate: {reqs_per_sec}/s | {mb_per_sec} MB/s]"
+            )
+
+    async def __aenter__(self):
+        await self._init_session()
+        return self
+
+    async def __aexit__(self, exc_type, exc_val, exc_tb):
+        await self._close()
+
+    async def _init_session(self):
+        if not self.session:
+            self.session = aiohttp.ClientSession(headers=self.headers)
+
+    async def _close(self):
+        if self.session:
+            await self.session.close()
+            self.session = None
+
+    async def _fetch_json(self, url):
+        """Fetch JSON with rate monitoring."""
+        async with self.limiter:
+            try:
+                url = fix_filing_url(url)
+                async with self.session.get(url) as response:
+                    if response.status == 429:
+                        raise RetryException(url)
+                    response.raise_for_status()
+                    content = await response.read()
+                    await self.rate_monitor.add_request(len(content))
+                    self.update_progress_description()
+                    return await response.json()
+            except aiohttp.ClientResponseError as e:
+                if e.status == 429:
+                    raise RetryException(url)
+                raise
+
+    async def _get_filing_urls_from_efts(self, base_url, submission_type=None):
+        """Fetch filing URLs from EFTS in batches with form type filtering."""
+        start = 0
+        page_size = 100
+        urls = []
+        
+        data = await self._fetch_json(f"{base_url}&from=0&size=1")
+        if not data or 'hits' not in data:
+            return []
+            
+        total_hits = data['hits']['total']['value']
+        if not total_hits:
+            return []
+
+        pbar = tqdm(total=total_hits, desc="Fetching URLs [Rate: 0/s | 0 MB/s]")
+        self.current_pbar = pbar
+        
+        while start < total_hits:
+            try:
+                tasks = [
+                    self._fetch_json(f"{base_url}&from={start + i * page_size}&size={page_size}") 
+                    for i in range(10)
+                ]
+                
+                results = await asyncio.gather(*tasks)
+                
+                for data in results:
+                    if data and 'hits' in data:
+                        hits = data['hits']['hits']
+                        if hits:
+                            # Filter hits based on exact form match
+                            if not submission_type or submission_type == "-0":
+                                filtered_hits = hits
+                            else:
+                                requested_forms = [submission_type] if isinstance(submission_type, str) else submission_type
+                                filtered_hits = [
+                                    hit for hit in hits
+                                    if hit['_source'].get('form', '') in requested_forms
+                                ]
+                            
+                            batch_urls = [
+                                f"https://www.sec.gov/Archives/edgar/data/{hit['_source']['ciks'][0]}/{hit['_id'].split(':')[0]}.txt" 
+                                for hit in filtered_hits
+                            ]
+                            urls.extend(batch_urls)
+                            pbar.update(len(hits))  # Update progress based on total hits processed
+                            self.update_progress_description()
+                
+                start += 10 * page_size
+
+            except RetryException as e:
+                print(f"\nRate limited. Sleeping for {e.retry_after} seconds...")
+                await asyncio.sleep(e.retry_after)
+                continue
+            except Exception as e:
+                print(f"\nError fetching URLs batch at {start}: {str(e)}")
+                break
+
+        pbar.close()
+        self.current_pbar = None
+        return urls
+        
+    async def _download_file(self, url, filepath):
+        """Download single file with precise rate limiting."""
+        async with self.connection_semaphore:
+            async with self.limiter:
+                try:
+                    url = fix_filing_url(url)
+                    async with self.session.get(url) as response:
+                        if response.status == 429:
+                            raise RetryException(url)
+                        response.raise_for_status()
+                        content = await response.read()
+                        await self.rate_monitor.add_request(len(content))
+                        self.update_progress_description()
+                        
+                        parsed_data = None
+                        if self.parse_filings:
+                            try:
+                                os.makedirs(os.path.dirname(filepath), exist_ok=True)
+                                async with aiofiles.open(filepath, 'wb') as f:
+                                    await f.write(content)
+
+                                parsed_data = parse_sgml_submission(
+                                    content=content.decode(), 
+                                    output_dir=os.path.dirname(filepath)
+                                )
+                                
+                                try:
+                                    os.remove(filepath)
+                                except Exception as e:
+                                    print(f"\nError deleting original file {filepath}: {str(e)}")
+                                    
+                            except Exception as e:
+                                print(f"\nError parsing {url}: {str(e)}")
+                                try:
+                                    os.remove(filepath)
+                                    parsed_dir = os.path.dirname(filepath) + f'/{url.split("/")[-1].split(".")[0].replace("-", "")}'
+                                    if os.path.exists(parsed_dir):
+                                        import shutil
+                                        shutil.rmtree(parsed_dir)
+                                except Exception as e:
+                                    print(f"\nError cleaning up files for {url}: {str(e)}")
+                        else:
+                            os.makedirs(os.path.dirname(filepath), exist_ok=True)
+                            async with aiofiles.open(filepath, 'wb') as f:
+                                await f.write(content)
+                        
+                        return filepath, parsed_data
+
+                except Exception as e:
+                    print(f"\nError downloading {url}: {str(e)}")
+                    return None
+
+    async def _download_worker(self, pbar):
+        """Worker to process download queue."""
+        while True:
+            try:
+                url, filepath = await self.download_queue.get()
+                result = await self._download_file(url, filepath)
+                if result:
+                    pbar.update(1)
+                self.download_queue.task_done()
+            except asyncio.CancelledError:
+                break
+            except Exception as e:
+                print(f"\nWorker error processing {url}: {str(e)}")
+                self.download_queue.task_done()
+
+    async def _download_and_process(self, urls, output_dir):
+        """Queue-based download processing."""
+        results = []
+        parsed_results = []
+        
+        pbar = tqdm(total=len(urls), desc="Downloading files [Rate: 0/s | 0 MB/s]")
+        self.current_pbar = pbar
+        
+        for url in urls:
+            filename = url.split('/')[-1]
+            filepath = os.path.join(output_dir, filename)
+            await self.download_queue.put((url, filepath))
+        
+        workers = [asyncio.create_task(self._download_worker(pbar)) 
+                  for _ in range(5)]  # Match number of workers to semaphore
+        
+        await self.download_queue.join()
+        
+        for worker in workers:
+            worker.cancel()
+        
+        await asyncio.gather(*workers, return_exceptions=True)
+        
+        pbar.close()
+        self.current_pbar = None
+        return results, parsed_results
+
+    def download_submissions(self, output_dir='filings', cik=None, ticker=None, submission_type=None, filing_date=None, parse=True):
+        """Main method to download SEC filings."""
+        self.parse_filings = parse
+        
+        async def _download():
+            async with self as downloader:
+                if ticker is not None:
+                    cik_value = identifier_to_cik(ticker)
+                else:
+                    cik_value = cik
+
+                params = {}
+                if cik_value:
+                    if isinstance(cik_value, list):
+                        params['ciks'] = ','.join(str(c).zfill(10) for c in cik_value)
+                    else:
+                        params['ciks'] = str(cik_value).zfill(10)
+
+                params['forms'] = ','.join(submission_type) if isinstance(submission_type, list) else submission_type if submission_type else "-0"
+
+                if isinstance(filing_date, list):
+                    dates = [(d, d) for d in filing_date]
+                elif isinstance(filing_date, tuple):
+                    dates = [filing_date]
+                else:
+                    date_str = filing_date if filing_date else f"2001-01-01,{datetime.now().strftime('%Y-%m-%d')}"
+                    start, end = date_str.split(',')
+                    dates = [(start, end)]
+
+                all_filepaths = []
+                all_parsed_data = []
+                
+                for start_date, end_date in dates:
+                    params['startdt'] = start_date
+                    params['enddt'] = end_date
+                    base_url = "https://efts.sec.gov/LATEST/search-index"
+                    efts_url = f"{base_url}?{urlencode(params, doseq=True)}"
+                    
+                    urls = await self._get_filing_urls_from_efts(efts_url,submission_type)
+                    if urls:
+                        filepaths, parsed_data = await self._download_and_process(urls, output_dir)
+                        all_filepaths.extend(filepaths)
+                        all_parsed_data.extend(parsed_data)
+
+                return all_filepaths, all_parsed_data
+
+        return asyncio.run(_download())
+
+    def download_company_concepts(self, output_dir='company_concepts', cik=None, ticker=None):
+        """Download company concept data."""
+        async def _download_concepts():
+            async with self as downloader:
+                if ticker is not None:
+                    ciks = identifier_to_cik(ticker)
+                elif cik:
+                    ciks = [cik] if not isinstance(cik, list) else cik
+                else:
+                    company_tickers = load_package_csv('company_tickers')
+                    ciks = [company['cik'] for company in company_tickers]
+
+                os.makedirs(output_dir, exist_ok=True)
+                urls = [f'https://data.sec.gov/api/xbrl/companyfacts/CIK{str(cik).zfill(10)}.json' for cik in ciks]
+                
+                pbar = tqdm(total=len(urls), desc="Downloading concepts [Rate: 0/s | 0 MB/s]")
+                self.current_pbar = pbar
+                
+                for url in urls:
+                    filename = url.split('/')[-1]
+                    filepath = os.path.join(output_dir, filename)
+                    await self.download_queue.put((url, filepath))
+                
+                workers = [asyncio.create_task(self._download_worker(pbar)) 
+                          for _ in range(5)]
+                
+                await self.download_queue.join()
+                
+                for worker in workers:
+                    worker.cancel()
+                
+                await asyncio.gather(*workers, return_exceptions=True)
+                
+                pbar.close()
+                self.current_pbar = None
+                
+                results = []
+                for url in urls:
+                    filename = url.split('/')[-1]
+                    filepath = os.path.join(output_dir, filename)
+                    if os.path.exists(filepath):
+                        results.append(filepath)
+                
+                return results
+
+        return asyncio.run(_download_concepts())