kernpy 0.0.2__py3-none-any.whl → 1.0.0__py3-none-any.whl

kernpy/polish_scores/download_polish_dataset.py

@@ -0,0 +1,357 @@
+import sys
+
+import requests
+from PIL import Image
+from io import BytesIO
+import os
+import json
+from typing import List, Optional
+
+from kernpy import ExportOptions, BEKERN_CATEGORIES, Importer, Exporter, Document, Encoding, read
+
+
+# This script creates the Polish dataset from the kern files.
+# It downloads both the systems and full pages
+DEFAULT_IIIF_ID = '/full/full/0/default.jpg'
+
+LOG_FILENAME = 'polish_index.json'
+
+
+def get_image_urls(_manifest_url):
+    # It returns the URL of pages tagged with a page number
+    # It returns a map with the page label as key, and the URL as
+    response = requests.get(_manifest_url)
+    manifest_data = response.json()
+    result = {}
+
+    # The corpus contains two kinds of IIIF manifests
+    if 'sequences' in manifest_data:
+        for sequence in manifest_data['sequences']:
+            for canvas in sequence['canvases']:
+                image_id = canvas['images'][0]['resource']['service']['@id']
+                page_label = canvas['label'] if 'label' in canvas else None
+                result[page_label] = image_id
+    else:
+        items = manifest_data.get('items', [])
+        for item in items:
+            pl = item.get('label').get('pl')
+            if pl:
+                page_label = pl[0]
+                if page_label != '[]':
+                    image_id = item.get('items')[0].get('items')[0].get('id', '')
+                    if image_id.endswith(DEFAULT_IIIF_ID):
+                        image_id = image_id[:-len(DEFAULT_IIIF_ID)]
+                    result[page_label] = image_id
+
+    # print(f'Items: ', len(items))
+
+    return result
+
+
+def download_and_save_image(url, save_path):
+    try:
+        # Make a GET request to the URL
+        response = requests.get(url)
+        response.raise_for_status()  # Raise an exception for bad responses
+
+        # Open the image using Pillow
+        image = Image.open(BytesIO(response.content))
+
+        # Save the image to the specified path
+        image.save(save_path, format='JPEG')
+
+        print(f"Image downloaded and saved to: {save_path}")
+    except Exception as e:
+        print(f"An error occurred: {e}")
+
+
+def factory_get_kern_type_exporter(kern_type: str) -> Encoding:
+    """
+    Factory method to get the Encoding
+
+    Args:
+        kern_type (str): the type of kern exporter. It can be 'krn' or 'ekrn'
+
+    Returns:
+        Encoding: the Encoding instance
+    """
+    if kern_type == 'krn':
+        return Encoding.normalizedKern
+    elif kern_type == 'ekrn':
+        return Encoding.eKern
+    else:
+        raise Exception(f'Unknown export kern type: {kern_type}')
+
+
+def extract_and_save_measures(document, from_measure, to_measure, krn_path, exporter_kern_type='ekrn'):
+    exporter_kern_type = factory_get_kern_type_exporter(exporter_kern_type)
+    export_options = ExportOptions(spine_types=['**kern'], token_categories=BEKERN_CATEGORIES, kern_type=exporter_kern_type)
+    export_options.from_measure = from_measure
+    export_options.to_measure = to_measure
+    exporter = Exporter()
+    exported_ekern = exporter.export_string(document, export_options)
+    with open(krn_path, "w") as f:
+        f.write(exported_ekern)
+
+
+def download_and_save_page_images(document, _output_path, map_page_label_iiif_ids, page_bounding_boxes, log_filename, exporter_kern_type='ekrn'):
+    print(f'Bounding boxes {page_bounding_boxes}')
+
+    for page_label, bounding_box_measure in page_bounding_boxes.items():
+        page_iiif_id = map_page_label_iiif_ids.get(page_label)
+        if page_iiif_id is None and page_label.startswith('#'):  # sometimes it's wrongly tagged without the #
+            page_iiif_id = map_page_label_iiif_ids.get(page_label[1:])
+
+        if page_iiif_id is not None:
+            bounding_box = bounding_box_measure.bounding_box
+            print(f"Page: {page_label}, "
+                  f"Bounding box: {bounding_box}, "
+                  f"ID: {page_iiif_id}, "
+                  f"from bar {bounding_box_measure.from_measure}, "
+                  f"to bar {bounding_box_measure.to_measure}")
+            url = os.path.join(page_iiif_id, bounding_box.xywh(), 'full', '0', 'default.jpg')
+            print(url)
+            image_path = os.path.join(_output_path, page_label + ".jpg")
+            download_and_save_image(url, image_path)
+            krn_path = os.path.join(_output_path, page_label + f'.{exporter_kern_type}')
+            extract_and_save_measures(document, bounding_box_measure.from_measure, bounding_box_measure.to_measure - 1,
+                                      krn_path, exporter_kern_type=exporter_kern_type)
+            add_log(document, krn_path, log_filename=log_filename)
+        else:
+            raise Exception(f'Cannot find IIIF id for page with label "{page_label}"')
+
+
+def findIIIFIds(document):
+    iiifTag = "!!!IIIF:"
+    for metacomment_token in document.get_metacomments():
+        encoding = metacomment_token
+        if encoding.startswith(iiifTag):
+            url = encoding[len(iiifTag):].strip()
+            print(f'Reading IIIF manifest from {url}')
+            return get_image_urls(url)
+    raise Exception('Cannot find any IIIF metacomment')
+
+
+def is_valid_document(document, kern_spines_filter) -> bool:
+    if kern_spines_filter is None:
+        return True
+
+    exporter = Exporter()
+    kern_types = exporter.get_spine_types(document, spine_types=['**kern'])
+    return len(kern_types) == int(kern_spines_filter)
+
+
+def convert_and_download_file(input_kern, _output_path, log_filename, kern_spines_filter: int = None, exporter_kern_type='ekrn') -> None:
+    document, errors = read(input_kern)
+    if len(errors) > 0:
+        print(f'ERRORS when kernpy.read:{input_kern} has errors {errors}\nContinue...', file=sys.stderr)
+        raise Exception(f'ERRORS when kernpy.read: {input_kern}. Has errors: {errors}')
+
+    if not is_valid_document(document, kern_spines_filter):
+        return
+
+    map_page_label_IIIF_ids = findIIIFIds(document)
+    download_and_save_page_images(document, _output_path, map_page_label_IIIF_ids, document.page_bounding_boxes,
+                                  log_filename=log_filename, exporter_kern_type=exporter_kern_type)
+
+
+def search_files_with_string(root_folder, target_string):
+    matching_files = []
+
+    for foldername, subfolders, filenames in os.walk(root_folder):
+        for filename in filenames:
+            if filename.endswith('.krn'):
+                file_path = os.path.join(foldername, filename)
+                try:
+                    with open(file_path, 'r', encoding='utf-8') as file:
+                        content = file.read()
+                        if target_string in content:
+                            relative_path = os.path.relpath(file_path, root_folder)
+                            matching_files.append(relative_path)
+                except Exception as e:
+                    print(f"Error reading {file_path}: {e}")
+
+    return matching_files
+
+
+def remove_extension(file_name):
+    # Using os.path.splitext to split the file name and extension
+    base_name, _ = os.path.splitext(file_name)
+    return base_name
+
+
+def add_log(document: Document, path, log_filename) -> None:
+    try:
+        def get_instruments(line):
+            words = line.split(' ')
+            instruments = []
+            for i in range(len(words)):
+                if words[i].isnumeric():
+                    instruments.extend([words[i + 1]] * int(words[i]))
+            return instruments
+
+        def get_publish_date(line):
+            if line is None or line == '':
+                return 0
+
+            clean_line = [char for char in line if char.isnumeric()]
+            return int(''.join(clean_line))
+
+        def round_publication_year(original_composer_date):
+            try:
+                if original_composer_date is None:
+                    return 0
+                start_date, end_date = original_composer_date.split('-')
+
+                start_year = int(start_date.split('/')[0])
+                end_year = int(end_date.split('/')[0])
+
+                RATIO = 0.7  # date where the composer was most active
+                return int(start_year + (end_year - start_year) * RATIO)
+            except Exception as e:
+                return -1
+
+        def round_publication_year_v2(original_composer_date):
+            def flatten(xss):
+                return [x for xs in xss for x in xs]
+
+            try:
+                items_date = original_composer_date.split('/')
+                clean_items = [item.replace(' ', '') for item in items_date]
+                clean_items = [item.replace('~', '') for item in clean_items]
+                split_again = [item.split('-') for item in clean_items]
+                flatten_items = flatten(split_again)
+                useful_items = [item for item in flatten_items if item.isnumeric()]
+                year_items = [int(item) for item in useful_items if len(item) == 4]
+                return int(year_items[0]) if len(year_items) > 0 else -3
+            except Exception as e:
+                return -2
+
+        info = {
+            'path': path,
+            'publication_date': get_publish_date(document.get_metacomments('PDT')[0]) if document.get_metacomments(
+                'PDT') else None,
+            'original_publication_date_tag': True,
+            'iiif': document.get_metacomments('IIIF')[0] if document.get_metacomments('IIIF') else None,
+            'n_measures': len(document.tree.stages),
+            'composer': document.get_metacomments('COM')[0] if document.get_metacomments('COM') else None,
+            'composer_dates': document.get_metacomments('CDT')[0] if document.get_metacomments('CDT') else None,
+            'tempo': document.get_metacomments('OTL')[0] if document.get_metacomments('OTL') else None,
+            'piece_title': document.get_metacomments('OPR')[0] if document.get_metacomments('OPR') else None,
+            'segment': document.get_metacomments('SEGMENT')[0] if document.get_metacomments('SEGMENT') else None,
+            'n_voices': len(get_instruments(document.get_metacomments('AIN')[0])) if document.get_metacomments(
+                'AIN') else 0,
+            'instruments': get_instruments(document.get_metacomments('AIN')[0]) if document.get_metacomments(
+                'AIN') else [],
+            'unique_instruments': [
+                *set(get_instruments(document.get_metacomments('AIN')[0]))] if document.get_metacomments('AIN') else [],
+        }
+
+        if info['publication_date'] in (0, 1, -1, -2) or info['publication_date'] is None:
+            info['publication_date'] = round_publication_year(info['composer_dates'])
+            info['original_publication_date_tag'] = False
+
+        if info['publication_date'] in (0, 1, -1, -2) or info['publication_date'] is None:
+            info['publication_date'] = round_publication_year_v2(info['composer_dates'])
+            info['original_publication_date_tag'] = False
+
+        with open(log_filename, 'a') as f:
+            json.dump(info, f)
+            f.write('\n')
+    except Exception as e:
+        print(f"Error adding log:{path}:{e}")
+
+
+def remove_empty_dirs(directory):
+    for root, dirs, files in os.walk(directory):
+        for dir in dirs:
+            full_dir = os.path.join(root, dir)
+            if not os.listdir(full_dir):
+                os.rmdir(full_dir)
+
+
+def store_error_log(filename, msg: dict):
+    with open(filename, encoding='utf-8', mode='a') as f:
+        f.write(f'{json.dumps(msg)}\n')
+
+
+def main(
+        input_directory: str,
+        output_directory: str,
+        remove_empty_directories: Optional[bool] = True,
+        kern_spines_filter: Optional[int] = 2,
+        exporter_kern_type: Optional[str] = 'ekern'
+) -> None:
+    """
+    Process the files in the input_directory and save the results in the output_directory.
+    http requests are made to download the images.
+
+    Args:
+        input_directory (str): directory where the input files are found
+        output_directory (str): directory where the output files are saved
+        remove_empty_directories (Optional[bool]): remove empty directories when finish processing the files
+        kern_spines_filter (Optional[int]): Only process files with the number of **kern spines specified.\
+            Use it to export 2-voice files. Default is 2.\
+            Use None to process all files.
+        exporter_kern_type (Optional[str]): the type of kern exporter. It can be 'krn' or 'ekrn'
+
+
+
+    Returns:
+        None
+
+    Examples:
+        >>> main('/kern_files', '/output_ekern')
+        None
+
+        >>> main('/kern_files', '/output_ekern', remove_empty_directories=False)
+        None
+
+        >>> main('/kern_files', '/output_ekern', kern_spines_filter=2, remove_empty_directories=False)
+        None
+
+        >>> main('/kern_files', '/output_ekern', kern_spines_filter=None, remove_empty_directories=False)
+        None
+
+        >>> main('/kern_files', '/output_ekern', exporter_kern_type='krn', remove_empty_directories=True)
+        None
+
+        >>> main('/kern_files', '/output_ekern', exporter_kern_type='ekrn', remove_empty_directories=True, kern_spines_filter=2)
+        None
+
+    """
+    print(f'Processing files in {input_directory} and saving to {output_directory}')
+    kern_with_bboxes = search_files_with_string(input_directory, 'xywh')
+    ok_files = []
+    ko_files = []
+    log_file = os.path.join(output_directory, LOG_FILENAME)
+    print(f"{25*'='}"
+          f"\nProcessing {len(kern_with_bboxes)} files."
+          f"\nLog will be saved in {log_file}."
+          f"\n{25*'='}")
+    for kern in kern_with_bboxes:
+        try:
+            filename = remove_extension(kern)
+            kern_path = os.path.join(input_directory, kern)
+            output_kern_path = os.path.join(output_directory, filename)
+            if not os.path.exists(output_kern_path):
+                os.makedirs(output_kern_path)
+            convert_and_download_file(kern_path, output_kern_path, log_filename=log_file, kern_spines_filter=kern_spines_filter, exporter_kern_type=exporter_kern_type)
+            ok_files.append(kern)
+        except Exception as error:
+            ko_files.append(kern)
+            print(f'Errors in {kern}: {error}')
+            store_error_log(os.path.join(output_directory, 'errors.json'), {'kern': kern, 'error': str(error)})
+
+    if remove_empty_directories:
+        remove_empty_dirs(output_directory)
+
+    print(f'----> OK files #{len(ok_files)}')
+    print(f'----> KO files #{len(ko_files)}')
+    print(ko_files)
+
+
+if __name__ == '__main__':
+    print(f'Usage: python -m kernpy --polish --input_directory /path/to/input --output_directory /path/to/output')
+    sys.exit(1)
+

kernpy/polish_scores/iiif.py

@@ -0,0 +1,47 @@
+import requests
+from PIL import Image
+from io import BytesIO
+import os
+
+
+def download_and_save_image(url, save_path):
+    try:
+        # Make a GET request to the URL
+        response = requests.get(url)
+        response.raise_for_status()  # Raise an exception for bad responses
+
+        # Open the image using Pillow
+        image = Image.open(BytesIO(response.content))
+
+        # Save the image to the specified path
+        image.save(save_path, format='JPEG')
+
+        print(f"Image downloaded and saved to: {save_path}")
+    except Exception as e:
+        print(f"An error occurred: {e}")
+
+
+def get_image_urls(_manifest_url):
+    # It returns the URL of pages tagged with a page number
+    # It returns a map with the page label as key, and the URL as
+    response = requests.get(_manifest_url)
+    manifest_data = response.json()
+
+    items = manifest_data.get('items', [])
+    print(f'Items: ', len(items))
+
+    for item in items:
+        pl = item.get('label').get('pl')
+        if pl:
+            page_num = pl[0]
+            if page_num != '[]':
+                url = item.get('items')[0].get('items')[0].get('id', '')
+                print(f'Page #{page_num}, {url}')
+                #image_path = os.path.join(output_folder, page_num + '.jpg')
+                #download_and_save_image(url, image_path)
+
+
+if __name__ == "__main__":
+    # Replace the manifest_url with the actual URL of your IIIF manifest
+    manifest_url = "https://polona2.pl/iiif/item/MTk4NjI5Mw/manifest.json"
+    get_image_urls(manifest_url)

kernpy/test_grammar.sh

@@ -0,0 +1,22 @@
+#!/bin/bash
+JAR=antlr-4.13.0-complete.jar
+TMP=/tmp/kerntest
+if [ $# -lt 1 ]
+then
+    echo "Use: <input files>"
+    exit 1
+fi
+
+mkdir ${TMP} 2> /dev/null
+cp kern/*g4 ${TMP}
+
+if [ ! -f ${TMP}/${JAR} ]; then
+  cp ${JAR} ${TMP}
+fi
+
+cd ${TMP}
+java -cp antlr-4.13.0-complete.jar org.antlr.v4.Tool kernLexer.g4
+java -cp antlr-4.13.0-complete.jar org.antlr.v4.Tool kernParser.g4
+javac -cp antlr-4.13.0-complete.jar *java
+cd -
+java -cp ${TMP}:antlr-4.13.0-complete.jar org.antlr.v4.gui.TestRig kern start $*

kernpy/util/__init__.py

@@ -0,0 +1,14 @@
+"""
+kernpy.util
+
+=====
+
+This module contains utility functions for the kernpy package.
+
+"""
+from .store_cache import *
+
+__all__ = [
+    'StoreCache'
+]
+

kernpy/util/helpers.py

@@ -0,0 +1,55 @@
+import warnings
+import functools
+
+
+def deprecated(reason: str):
+    """
+    Decorator to mark a function or class as deprecated.
+    
+    Args:
+        reason (str): The reason why the function/class is deprecated.
+    
+    Example:
+        @deprecated("Use new_function instead.")
+        def old_function():
+            pass
+    """
+    def decorator(func):
+        @functools.wraps(func)
+        def wrapper(*args, **kwargs):
+            warnings.warn(
+                f"'{func.__name__}' is deprecated: {reason}",
+                category=DeprecationWarning,
+                stacklevel=2,
+            )
+            return func(*args, **kwargs)
+        return wrapper
+    return decorator
+
+
+def deprecated_argument(arg_name: str, reason: str):
+    """
+    Decorator to mark a specific argument as deprecated.
+    
+    Args:
+        arg_name (str): The deprecated argument name.
+        reason (str): The reason why the argument is deprecated.
+    
+    Example:
+        @deprecated_argument("old_param", "Use 'new_param' instead.")
+        def my_function(new_param, old_param=None):
+            pass
+    """
+    def decorator(func):
+        @functools.wraps(func)
+        def wrapper(*args, **kwargs):
+            if arg_name in kwargs:
+                warnings.warn(
+                    f"Argument '{arg_name}' in '{func.__name__}' is deprecated: {reason}",
+                    category=DeprecationWarning,
+                    stacklevel=2,
+                )
+            return func(*args, **kwargs)
+        return wrapper
+    return decorator
+

kernpy/util/store_cache.py

@@ -0,0 +1,35 @@
+class StoreCache:
+    """
+    A simple cache that stores the result of a callback function
+    """
+    def __init__(self):
+        """
+        Constructor
+        """
+        self.memory = {}
+
+    def request(self, callback, request):
+        """
+        Request a value from the cache. If the value is not in the cache, it will be calculated by the callback function
+        Args:
+            callback (function): The callback function that will be called to calculate the value
+            request (any): The request that will be passed to the callback function
+
+        Returns (any): The value that was requested
+
+        Examples:
+            >>> def add_five(x):
+            ...     return x + 5
+            >>> store_cache = StoreCache()
+            >>> store_cache.request(callback, 5)  # Call the callback function
+            10
+            >>> store_cache.request(callback, 5)  # Return the value from the cache, without calling the callback function
+            10
+        """
+        if request in self.memory:
+            return self.memory[request]
+        else:
+            result = callback(request)
+            self.memory[request] = result
+            return result
+

kernpy/visualize_analysis.sh

@@ -0,0 +1,23 @@
+#!/bin/bash
+JAR=antlr-4.13.0-complete.jar
+TMP=/tmp/kerntest
+if [ $# -lt 1 ]
+then
+    echo "Use: <input file>"
+    exit 1
+fi
+
+
+mkdir ${TMP} 2> /dev/null
+cp kern/*g4 ${TMP}
+
+if [ ! -f ${TMP}/${JAR} ]; then
+  cp ${JAR} ${TMP}
+fi
+
+cd ${TMP}
+java -cp antlr-4.13.0-complete.jar org.antlr.v4.Tool kernLexer.g4
+java -cp antlr-4.13.0-complete.jar org.antlr.v4.Tool kernParser.g4
+javac -cp antlr-4.13.0-complete.jar *java
+cd -
+java -cp ${TMP}:antlr-4.13.0-complete.jar org.antlr.v4.gui.TestRig kern start -gui $1