diff --git a/README.md b/README.md
deleted file mode 100644
index 9635f915e6b6283abb229bee1ae84b46afbe734b..0000000000000000000000000000000000000000
--- a/README.md
+++ /dev/null
@@ -1,3 +0,0 @@
-# Projekt CiS-Biochemie 2021-22
-
-Automated citation loop finder for ACS articles.
\ No newline at end of file
diff --git a/citation_parser_ui.py b/citation_parser_ui.py
new file mode 100644
index 0000000000000000000000000000000000000000..66529c37a74c6416c34d296246a25c61761ef0cb
--- /dev/null
+++ b/citation_parser_ui.py
@@ -0,0 +1,205 @@
+import base64
+import re
+import dash
+from dash import dcc
+from dash import html
+from dash import callback_context
+from dash.dependencies import Input, Output, State
+from dash.exceptions import PreventUpdate
+from input.interface import InputInterface
+import input.publication
+
+app = dash.Dash(__name__)
+
+additional_options = ['Update Automatically','Smart Input']
+
+app.layout = html.Div([
+ # Layer 0: For the Header and Help Function(s)
+ html.Div([
+ html.Button(id='show-info',children='Show Info',n_clicks=0),
+ html.Div(id='info-box')
+ ]),
+ # Layer 1: For all mandatory Inputs
+ html.Div([
+ "Input: ",
+ dcc.Input(id='input-string', value='', type='text',debounce=True),
+ dcc.Input(id='forward-depth',value='1',type='number',min='1',max='10'),
+ dcc.Input(id='backward-depth',value='1',type='number',min='1',max='10'),
+ dcc.Upload(
+ id="upload-data",
+ children=html.Div(
+ ["Drag and drop or click to select a file to upload."]
+ ),
+
+ style={
+ "width": "30%",
+ "height": "60px",
+ "lineHeight": "60px",
+ "borderWidth": "1px",
+ "borderStyle": "dashed",
+ "borderRadius": "5px",
+ "textAlign": "center",
+ "margin": "10px",
+ })
+ ]),
+ # Layer 2: For the checklist, Remove-/Start-Buttons and input-error-message
+ html.Div([
+ dcc.Checklist(id='input-checklist',options=[],labelStyle = dict(display='block'),value=[]),
+ html.Div(id='input-err',style={'color':'red'}),
+ html.Button(id='clear-all-button',children='Clear All'),
+ html.Button(id='clear-selected-button',children='Clear Selected'),
+ html.Button(id='start-button',children='Generate Graph')
+ ]),
+ # Layer 3: For additional Options (e.g. Topological Sort)
+ html.Div([
+ html.H4('Additional Options'),
+ dcc.Checklist(id='additional-options',
+ options=[{'label':k,'value':k} for k in additional_options],
+ value=[])
+ ]),
+ # Layer 4: For the Graph
+ html.Div([
+ html.Div(id='test-output')
+ ])
+])
+
+'''
+Most important callback function. Updates the checklist that holds all inputs.
+input-string is required as Output to clear the input box after each input
+'''
+@app.callback(
+ Output('input-checklist','options'),
+ Output('input-checklist','value'),
+ Output('input-string','value'),
+ Output('input-err','children'),
+ Input('input-string','value'),
+ Input('clear-all-button','n_clicks'),
+ Input('clear-selected-button','n_clicks'),
+ Input('upload-data','contents'),
+ State('input-checklist','options'),
+ State('input-checklist','value'),
+ State('additional-options','value')
+)
+def update_input_checklist(input_value,btn1,btn2,filecontents,all_inputs,selected_inputs,additional_options):
+ '''
+ :param input_value: given by dcc.Input
+ :type input_value: string
+ :param btn1: signals pressing of clear-all-button
+ :param btn2: signals pressing of clear-selected-button
+ :param all_inputs: all labels and values from the checklist,
+ regardless if they have been checked or not
+ :type all_inputs: list of dictionaries with 2 entries each
+ :param selected_inputs: values of all checked elements
+ :type selected_inputs: list of strings
+ '''
+ changed_id = [p['prop_id'] for p in callback_context.triggered][0]
+ # if clear-all-button was pressed:
+ if 'clear-all-button' in changed_id:
+ return list(),list(),'',''
+ # if clear-selected-button was pressed:
+ if 'clear-selected-button' in changed_id:
+ all_inputs = [i for i in all_inputs if i['value'] not in selected_inputs]
+ return all_inputs,list(),'',''
+ # when a new element is added via dcc.Input
+ if 'input-string' in changed_id:
+ options = all_inputs
+ currValues = [x['value'] for x in options]
+ if input_value not in currValues:
+ if 'Smart Input' in additional_options:
+ try:
+ i = InputInterface()
+ pub = i.get_publication(input_value)
+ except Exception as err:
+ return options,selected_inputs,'','{}'.format(err)
+ rep_str = pub.contributors[0] + ',' + pub.journal + ',' + pub.publication_date
+ options.append({'label':rep_str, 'value':input_value})
+ else:
+ options.append({'label':input_value,'value':input_value})
+ return options,selected_inputs,'',''
+ # when a txt-file is uploaded
+ if 'upload-data.contents' in changed_id:
+ if filecontents:
+ string = filecontents
+ found = base64.b64decode(re.search(',(.+?)$', string).group(1))
+ filecontents = found.decode('utf-8')
+ list_of_inputs = (filecontents.strip().split('\n'))
+ options = all_inputs
+ CurrValues = [x['value'] for x in options]
+ for input_value in list_of_inputs:
+ if input_value not in CurrValues:
+ if 'Smart Input' in additional_options:
+ try:
+ i = InputInterface()
+ pub = i.get_publication(input_value)
+ except Exception as err:
+ return options,selected_inputs,'','{}'.format(err)
+ rep_str = pub.contributors[0] + ',' + pub.journal + ',' + pub.publication_date
+ options.append({'label':rep_str, 'value':input_value})
+ else:
+ options.append({'label':input_value,'value':input_value})
+ return options,selected_inputs,'',''
+ # when the programm is first started:
+ if input_value == '':
+ app.layout['input-checklist'].options.clear()
+ return list(),list(),'',''
+
+'''
+This callback shows and hides the (first) help-box
+'''
+@app.callback(
+ Output('info-box','children'),
+ Input('show-info','n_clicks')
+)
+def show_hide_info_box(n_clicks):
+ if n_clicks % 2 == 0:
+ return ''
+ else:
+ return 'Hier koennte Ihre Werbung stehen'
+
+'''
+Basic structure for a callback that generates an output
+'''
+@app.callback(
+ Output('test-output','children'),
+ Input('start-button','n_clicks'),
+ Input('input-checklist','options'),
+ Input('input-checklist','value'),
+ Input('forward-depth','value'),
+ Input('backward-depth','value'),
+ State('additional-options','value')
+)
+def generate_output(n_clicks,all_inputs,selected_inputs,
+ forward_depth,backward_depth,additional_options):
+ '''
+ :param n_clicks: how often has Generate Graph been clicked
+ :type n_clicks: int
+ :param all_inputs: all labels and values from the checklist,
+ regardless if they have been checked or not
+ :type all_inputs: list of dictionaries with 2 entries each
+ :param selected_inputs: values of all checked elements
+ :type selected_inputs: list of strings
+ :param forward_depth: forward recursion depth
+ :type forward_depth: unsigned int
+ :param backward_depth: backward recursion depth
+ :type backward_depth: unsigned int
+ :param additional_options: value of all selected additional options
+ :type additional_options: list of strings
+ '''
+ changed_id = [p['prop_id'] for p in callback_context.triggered][0]
+ if n_clicks is None:
+ raise PreventUpdate
+ elif 'Update Automatically' in additional_options \
+ or 'start-button' in changed_id:
+ s = ''
+ for i in range(len(all_inputs)):
+ x = all_inputs[i]['value']
+ if x in selected_inputs:
+ s += x*(abs(int(forward_depth)-int(backward_depth)))
+ else:
+ s += x*(int(forward_depth)+int(backward_depth))
+ return s
+ else:
+ raise PreventUpdate
+
+if __name__ == '__main__':
+ app.run_server(debug=True)
diff --git a/example_urls b/example_urls
new file mode 100644
index 0000000000000000000000000000000000000000..96ac680c65edddcb495312000157edea1ab94884
--- /dev/null
+++ b/example_urls
@@ -0,0 +1,2 @@
+https://pubs.acs.org/doi/10.1021/acs.jcim.5b00332
+https://pubs.acs.org/doi/10.1021/acs.jcim.6b00709
diff --git a/input/README.md b/input/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..110ce69136a8935b83d070113130222f243e924f
--- /dev/null
+++ b/input/README.md
@@ -0,0 +1,50 @@
+# Projekt CiS-Projekt 2021/22
+
+Input-Package to fetch publication information with a given url.
+
+## Usage/Examples
+
+```python
+from input.interface import InputInterface as Input
+from input.publication import Publication
+
+def main(url):
+ inter = Input()
+ try:
+ pub = inter.get_publication(url)
+ except Exception as error:
+ raise error
+
+ print(pub)
+ pub.title = "Cool new Title"
+ print(pub)
+
+if __name__ == "__main__":
+ main("https://doi.org/10.1021/acs.chemrev.8b00728")
+```
+
+The expected results of calling this methode are:
+| Input-Url | Result |
+|-----------|-----------|
+| supported & correct| A publication Instance |
+| supported & uncorrect| ValueError|
+| not supported | ValueError|
+
+Supported Url are urls, which comply with the url-pattern of supported Journals.
+
+### Supported Journals:
+
+- ACS-Journals
+- (Nature-Journals)
+
+## Testing
+
+``` c
+python -m unittest input/test/<file.py> -v
+# for all tests in directory
+python -m unittest discover input/test -v
+```
+## Authors
+- Florian Jochens
+- Sam Ockenden
+- Julius Schenk
\ No newline at end of file
diff --git a/input/__init__.py b/input/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/input/__pycache__/__init__.cpython-38.pyc b/input/__pycache__/__init__.cpython-38.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..dd915a8709319a33f27a6083fd415baeb6b4d6c3
Binary files /dev/null and b/input/__pycache__/__init__.cpython-38.pyc differ
diff --git a/input/__pycache__/interface.cpython-38.pyc b/input/__pycache__/interface.cpython-38.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..b823637b2913428dc68e9eb088266aafd5643971
Binary files /dev/null and b/input/__pycache__/interface.cpython-38.pyc differ
diff --git a/input/__pycache__/publication.cpython-38.pyc b/input/__pycache__/publication.cpython-38.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..b893491e387cafe8f592479d23c76f302b721bde
Binary files /dev/null and b/input/__pycache__/publication.cpython-38.pyc differ
diff --git a/input/get/__init__.py b/input/get/__init__.py
new file mode 100755
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/input/get/__pycache__/__init__.cpython-38.pyc b/input/get/__pycache__/__init__.cpython-38.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..8f81f36aaccfa12fdee27996811a2c67cc19fff8
Binary files /dev/null and b/input/get/__pycache__/__init__.cpython-38.pyc differ
diff --git a/input/get/__pycache__/acs.cpython-38.pyc b/input/get/__pycache__/acs.cpython-38.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..0f8c362e84d5ccbbdf32f8310eeafd1376ab1977
Binary files /dev/null and b/input/get/__pycache__/acs.cpython-38.pyc differ
diff --git a/input/get/__pycache__/journal_fetcher.cpython-38.pyc b/input/get/__pycache__/journal_fetcher.cpython-38.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..894af00c5d09c44d55f9e5d9a60d05409a7620db
Binary files /dev/null and b/input/get/__pycache__/journal_fetcher.cpython-38.pyc differ
diff --git a/input/get/__pycache__/nature.cpython-38.pyc b/input/get/__pycache__/nature.cpython-38.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..47affbf4e197c535229a9e2bbecf3988abfad115
Binary files /dev/null and b/input/get/__pycache__/nature.cpython-38.pyc differ
diff --git a/input/get/acs.py b/input/get/acs.py
new file mode 100755
index 0000000000000000000000000000000000000000..9691845b27ae694a8213a0f0fe5f827c75890eee
--- /dev/null
+++ b/input/get/acs.py
@@ -0,0 +1,192 @@
+#!/usr/bin/env python3
+
+"""
+Child class of JournalFetcher
+Usage: Check if Url can be used with 'can_use_url'
+ and then fetch publication with 'get_publication'
+"""
+
+import re
+
+from input.get.journal_fetcher import JournalFetcher
+from input.publication import Publication, Citation
+
+
+class Fetcher(JournalFetcher):
+ """
+ Specific Fetcher for the ACS journals.
+ """
+
+ # Constant for the abbreviations of the supported Journals
+ SUPPORTED_JOURNALS = ['1021']
+
+ @staticmethod
+ def can_use_url(url: str) -> str:
+ """
+ Uses Regex to extract journal specific substrings in Doi.
+ TODO: Support non Doi-urls
+ """
+ matched_url = re.match(r'^(https?://)?(doi.org/|pubs.acs.org/doi/)?(10.(\d{4})/\w+.\S+)', url.strip(". \t\r\n"))
+
+ #Checks if match exists
+ if matched_url is not None:
+ return matched_url[4] in Fetcher.SUPPORTED_JOURNALS
+ else:
+ return False
+
+ @staticmethod
+
+
+ def get_pub_light(url: str) -> Publication:
+ """
+ Fetches html and creates Beatifulsoup-instance in parent class.
+ Specific css-searches for ACS-Journals and creates Publication-instance.
+ """
+
+ # Creation of Soup
+ try:
+ soup = JournalFetcher.get_soup(url)
+ except Exception as error:
+ raise error
+
+ # Raise Error if re recognizes Pattern, but url isnt correct:
+ # For other Urls
+ if soup.text.strip(" \t\n")=="Missing resource null":
+ raise ValueError("'{}' matches Pattern for 'ACS', but doesnt link to Paper.".format(url))
+
+ # For Dois
+ if soup.title is not None:
+ if soup.title.text == "Error: DOI Not Found":
+ raise ValueError("'{}' matches Pattern for 'ACS', but doesnt link to Paper.".format(url))
+
+
+ soup_header = soup.select('.article_header')[0]
+
+ # Creates Publication
+ doi_url = soup_header.select('a[title="DOI URL"]')[0].string
+ title = soup_header.select(".hlFld-Title")[0].text
+
+ contributors = []
+ for author in soup_header.select(".hlFld-ContribAuthor"):
+ contributors.append(author.text)
+
+ journal = soup_header.select(".cit-title")[0].text
+
+ # Replaces abbreviation with whole name
+ if journal in JournalFetcher.abbrev_dict:
+ journal = JournalFetcher.abbrev_dict[journal]
+
+
+ published = soup_header.select(".pub-date-value")[0].text
+
+ subjects = []
+ subject_soup = soup_header.select('.article_header-taxonomy')[0]
+ for subject in subject_soup.select('a'):
+ subjects.append(subject.text)
+
+ return Publication(doi_url, title, contributors, journal, published,
+ subjects)
+
+ def get_publication(url: str) -> Publication:
+ """
+ Fetches html and creates Beatifulsoup-instance in parent class.
+ Specific css-searches for ACS-Journals and creates Publication-instance.
+ """
+
+ # Creation of Soup
+ try:
+ soup = JournalFetcher.get_soup(url)
+ except Exception as error:
+ raise error
+
+ # Raise Error if re recognizes Pattern, but url isnt correct:
+ # For other Urls
+ if soup.text.strip(" \t\n")=="Missing resource null":
+ raise ValueError("'{}' matches Pattern for 'ACS', but doesnt link to Paper.".format(url))
+
+ # For Dois
+ if soup.title is not None:
+ if soup.title.text == "Error: DOI Not Found":
+ raise ValueError("'{}' matches Pattern for 'ACS', but doesnt link to Paper.".format(url))
+
+
+ soup_header = soup.select('.article_header')[0]
+
+ #Could be used for more specific search
+ ref_cit_soup = soup
+
+ # Creates Publication
+ doi_url = soup_header.select('a[title="DOI URL"]')[0].string
+ title = soup_header.select(".hlFld-Title")[0].text
+
+ contributors = []
+ for author in soup_header.select(".hlFld-ContribAuthor"):
+ contributors.append(author.text)
+
+ journal = soup_header.select(".cit-title")[0].text
+
+ # Replaces abbreviation with whole name
+ if journal in JournalFetcher.abbrev_dict:
+ journal = JournalFetcher.abbrev_dict[journal]
+
+
+ published = soup_header.select(".pub-date-value")[0].text
+
+ subjects = []
+ subject_soup = soup_header.select('.article_header-taxonomy')[0]
+ for subject in subject_soup.select('a'):
+ subjects.append(subject.text)
+
+
+ references = []
+ references_soup = ref_cit_soup.select('ol#references')
+ if references_soup != []:
+ for reference in references_soup[0].select('li'):
+ if reference.select('.refDoi') != []:
+ ref_doi = "https://doi.org/{}".format(reference.select('.refDoi')[0].text.strip()[5:])
+ else:
+ # No Doi -> No Paper
+ continue
+ ref_title = reference.select('.NLM_article-title')[0].text\
+ if reference.select('.NLM_article-title') != [] else None
+ ref_journal = reference.select('i')[0].text\
+ if reference.select('i') != [] else None
+
+ # Replaces abbreviation with whole name
+ if ref_journal in JournalFetcher.abbrev_dict:
+ ref_journal = JournalFetcher.abbrev_dict[ref_journal]
+
+ ref_contributors=[]
+ for author in reference.select('.NLM_contrib-group'):
+ ref_contributors.append(author.text.replace("\n", " ").replace("\r", ""))
+
+ references.append(Citation(ref_doi, ref_title, ref_journal, ref_contributors, cit_type="Reference"))
+
+ citations = []
+ citation_soup = ref_cit_soup.select('.cited-content_cbyCitation')
+ if citation_soup != []:
+ for citation in citation_soup[0].select('li'):
+ if citation.select('a[title="DOI URL"]') != []:
+ cit_doi = citation.select('a[title="DOI URL"]')[0].text
+ else:
+ # No Doi -> No Paper
+ continue
+ cit_title = citation.select('.cited-content_cbyCitation_article-title')[0].text\
+ if citation.select('.cited-content_cbyCitation_article-title')!= [] else None
+ cit_journal = citation.select('.cited-content_cbyCitation_journal-name')[0].text\
+ if citation.select('.cited-content_cbyCitation_journal-name') != [] else None
+
+ # Replaces abbreviation with whole name
+ if cit_journal in JournalFetcher.abbrev_dict:
+ cit_journal = JournalFetcher.abbrev_dict[cit_journal]
+ cit_contributors =[]
+ cit_contributors = citation.select('.cited-content_cbyCitation_article-contributors')[0]\
+ .text.replace("\n", " ").replace("\r", "").split(', ')
+ # clean up of the last Entry
+ cit_contributors_last = cit_contributors.pop().strip(". ")
+ if cit_contributors_last != '':
+ cit_contributors.append(cit_contributors_last)
+ citations.append(Citation(cit_doi, cit_title, cit_journal, cit_contributors, cit_type = "Citation"))
+
+ return Publication(doi_url, title, contributors, journal, published
+ , subjects, references, citations)
diff --git a/input/get/journal_fetcher.py b/input/get/journal_fetcher.py
new file mode 100755
index 0000000000000000000000000000000000000000..514af1f80f5c7d442b790aebf5fe3954d50f8f5d
--- /dev/null
+++ b/input/get/journal_fetcher.py
@@ -0,0 +1,96 @@
+#!/usr/bin/env python3
+
+"""
+Parent class for specific Journal
+"""
+
+from abc import ABCMeta, abstractmethod
+from bs4 import BeautifulSoup
+import requests
+from input.publication import Publication
+
+
+class JournalFetcher(metaclass=ABCMeta):
+ """
+ This is a abstract-class for fetcher modules
+ """
+
+ @staticmethod
+ def get_soup(url: str) -> BeautifulSoup:
+ """
+ Retrieves webside-html and returns a BeautifulSoup-instance
+
+ Parameters:
+ -----------
+ :type url: str
+ :param url: doi-url to a publication
+ :return: BeatifulSoup-instance
+ """
+ try:
+ req = requests.get(url)
+ except requests.exceptions.HTTPError as err:
+ raise SystemExit(err)
+
+ return BeautifulSoup(req.content, 'html.parser')
+
+
+ @staticmethod
+ @abstractmethod
+ def can_use_url(url: str) -> bool:
+ """
+ Abstract-function to be implemented in subclass.
+ Checks if given url links to a supported journal
+ """
+ raise AttributeError("JournalFetcher for '{}' hasnt implemented 'can_use_url()'".format(url))
+
+
+ @staticmethod
+ @abstractmethod
+ def get_publication(url: str) -> Publication:
+ """
+ Abstract-function to be implemented in subclass.
+ Creates a Publication-instance.
+ """
+ raise AttributeError("JournalFetcher for '{}' hasnt implemented 'get_publication()'".format(url))
+
+
+ # A Dictionary, which connects abbreviation to whole journal-name
+ abbrev_dict = {
+ "Nat. Protoc.":"Journal of Natural Products"
+ ,"PLoS Comput. Biol.":"PLoS Computational Biology"
+ ,"PLoS One":"PLoS One"
+ ,"Protein Sci.":"Protein Science"
+ ,"J. Am. Chem. Soc.":"Journal of the American Chemical Society"
+ ,"J. Chem. Phys.":"Journal of Chemical Physics"
+ ,"Appl. Sci.":"Applied Science"
+ ,"Comput. Sci. Eng.":"Computing in Science & Engineering"
+ ,"Beilstein J. Org. Chem.":"Beilstein Journal of Organic Chemistry"
+ ,"Biol. Chem.":"Biological Chemistry"
+ ,"Isr. J. Chem.":"Israel Journal of Chemistry"
+ ,"Nat. Methods":"Nature Methods"
+ ,"Proc. Natl. Acad. Sci. U. S. A.":"Proceedings of the National Academy of Sciences of the United States of America"
+ ,"J. Phys. Chem. B":"Journal of Physical Chemistry B"
+ ,"Carbohydr. Res.":"Carbohydrate Research"
+ ,"J. Chem. Theory Comput.":"Journal of Chemical Theory and Computation"
+ ,"J. Mol. Biol.":"Journal of Molecular Biology"
+ ,"Nucleic Acids Res.":"Nucleic Acids Research"
+ ,"J. Comput. Chem.":"Journal of Computational Chemistry"
+ ,"J. Cheminf.":"Journal of Cheminformatics"
+ ,"J. Med. Chem.":"Journal of Medicinal Chemistry"
+ ,"J. Comput.-Aided Mol. Des.":"Journal of Computer-Aided Molecular Design"
+ ,"J. Chem. Inf. Model.":"Journal of Chemical Information and Modeling"
+ ,"Mol. Cell":"Molecular Cell"
+ ,"J. Cell Biolog.":"Journal of Cell Biology"
+ ,"Mol. Cell Biol.":"Molecular and Cellular Biology"
+ ,"J. Cell Sci.":"Journal of Cell Science"
+ ,"Nat. Cell Biol.":"Nature Cell Biology"
+ ,"J. Aerosol Sci. Technol.":"Aerosol Science and Technology"
+ ,"Mol. Biol. Cell":"Molecular Biology of the Cell"
+ ,"Build. Environ.":"Building and Environment"
+ ,"Sci. Rep.":"Scientific Reports"
+ ,"Nat. Chem.":"Nature Chemistry"
+ ,"Nat. Med.":"Nature Medicine"
+ ,"Nat. Commun.":"Nature Communications"
+ ,"Exp. Cell Res.":"Experimental Cell Research"
+ ,"Nat. Chem. Biol.":"Nature Chemical Biology"
+ }
\ No newline at end of file
diff --git a/input/get/nature.py b/input/get/nature.py
new file mode 100644
index 0000000000000000000000000000000000000000..c50ea0ef9d1d4a9a386730e31cc72372cbf698c0
--- /dev/null
+++ b/input/get/nature.py
@@ -0,0 +1,59 @@
+#!/usr/bin/env python3
+
+"""
+Child class of JournalFetcher
+Usage: Check if Url can be used with 'can_use_url'
+ and then fetch publication with 'get_publication'
+"""
+
+# import re
+from input.get.journal_fetcher import JournalFetcher
+from input.publication import Publication
+
+
+class Fetcher(JournalFetcher):
+
+ """
+ scrapes publication metadata from a provided url
+ """
+
+ # TODO: List of Compatable Journals
+ # NOTE: nature does not use journal names in doi links, must match by 10.xxxx identifier instead
+ SUPPORTED_JOURNALS = []
+
+ @staticmethod
+ def can_use_url(url: str) -> bool:
+ """
+ Checks if given url links to a supported journal.
+ """
+
+ # TODO: Check the URL for compatability
+ # re.match in SUPPORTED_JOURNALS
+ return False
+
+ @staticmethod
+ def get_publication(url: str) -> Publication:
+ """
+ Creates a Publication-instance.
+ """
+
+ soup = JournalFetcher.get_soup(url)
+
+ _doi_url = "https://doi.org/" + soup.head.find(attrs={"name": "DOI"}).get("content")
+ _title = soup.head.find(attrs={"name": "citation_title"}).get("content")
+ _journal = soup.head.find(attrs={"name": "citation_journal_title"}).get("content")
+ _published = soup.head.find(attrs={"name": "prism.publicationDate"}).get("content")
+ _contributors = []
+ _subjects = []
+
+ for creator in soup.head.findAll(attrs={"name": "dc.creator"}):
+ _contributors.append(creator.get("content"))
+
+ for subject in soup.head.findAll(attrs={"name": "dc.subject"}):
+ _subjects.append(subject.get("content"))
+
+ return Publication(_doi_url, _title, _contributors, _journal, _published, _subjects)
+
+ # TODO: Exceptions-handling
+ # raise ValueException("Cant Fetch: '{}'".format(error))
+ # return None
diff --git a/input/get/template_.py b/input/get/template_.py
new file mode 100755
index 0000000000000000000000000000000000000000..58de0237bd514f7dd1b5b25f251b740d33e3589e
--- /dev/null
+++ b/input/get/template_.py
@@ -0,0 +1,51 @@
+#!/usr/bin/env python3
+
+"""
+Child class of JournalFetcher
+Usage: None, this is just a template and should be ignored
+"""
+
+# import re
+from input.get.journal_fetcher import JournalFetcher
+from input.publication import Publication
+
+
+class Fetcher(JournalFetcher):
+
+ """
+ This is only a template and therefore has no functionality
+ """
+
+ # TODO: Naming-Convention:
+ # Class: 'Fetcher'
+ # file: [journal-/organisation-name]
+ # format = "[a-z]*.py" allowed
+ # TODO: List of Compatable Journals
+ SUPPORTED_JOURNALS = []
+
+ @staticmethod
+ def can_use_url(url: str) -> bool:
+ """
+ Checks if given url links to a supported journal.
+ """
+
+ # TODO: Check the URL for compatability
+ # url_re = re.match(r'(https?://)?(doi.org/)?(10.(\d{4})/\w+.\S+)', url)
+ # if url_re is not None:
+ # return url_re[4] in SUPPORTED_JOURNALS
+ # else:
+ return False
+
+ @staticmethod
+ def get_publication(url: str) -> Publication:
+ """
+ Creates a Publication-instance.
+ """
+
+ # TODO: Fetch data from the HTML
+ # soup = JournalFetcher.get_soup(url)
+ # doi,title,contributors[],journal,publication_date,subjects[],references[],citations[]
+ # TODO: Create new Publication-instance
+ # return Publication(doi_url, title, contributors = [], journal
+ # , publication_date, subjects = [], references = [], citations = [])
+ return None
\ No newline at end of file
diff --git a/input/interface.py b/input/interface.py
new file mode 100755
index 0000000000000000000000000000000000000000..59515b3a3a2a5361222b8e55d3a7314ab3907132
--- /dev/null
+++ b/input/interface.py
@@ -0,0 +1,113 @@
+#!/usr/bin/env python3
+
+"""
+Interface for the Input-Package only this should be accessed from outside this Package.
+
+"""
+from os import walk
+import importlib
+import pathlib
+import re
+from input.publication import Publication
+
+class InputInterface:
+ """
+ Singleton which dynamically imports and manages fetchers
+ """
+
+ instance = None
+ get_path = None
+ fetcher_classes=[]
+
+ # '__new__' is called before '__init__' and gives us an instance
+ def __new__(cls, *args, **kwargs):
+
+ # checks if an instance exists and if it doesnt creates one
+ if cls.instance == None:
+ cls.instance = super(InputInterface, cls).__new__(cls,*args, **kwargs)
+
+ return cls.instance
+
+ def __init__(self):
+ # imports all modules
+
+ if self.fetcher_classes ==[]:
+ self.import_fetcher_classes()
+ if self.fetcher_classes ==[]:
+ raise AttributeError("No specific Fetchers where found at: '{}'"
+ .format(self.get_path))
+
+
+ def get_publication(self, url: str) -> Publication:
+ """
+ The interface-method to get a Publication-instance
+ (including it's citations and references)
+
+ Parameters
+ ----------
+ :param url: url to a Publication
+ :type url: str
+ :return: Publication instance or None if not supported
+ """
+
+ # Checks if module supports the 'url' and
+ # returns a Publication if it does.
+ for fetcher_class in InputInterface.fetcher_classes:
+ if fetcher_class.can_use_url(url):
+ return fetcher_class.get_publication(url)
+
+ # No Module for given url was found
+ raise ValueError("'{}' is not supported".format(url))
+
+ def get_pub_light(self, url: str) -> Publication:
+ """
+ The interface-method to get a Publication-instance
+ (only for main article)
+
+ Parameters
+ ----------
+ :param url: url to a Publication
+ :type url: str
+ :return: Publication instance or None if not supported
+ """
+
+ # Checks if module supports the 'url' and
+ # returns a Publication if it does.
+ for fetcher_class in InputInterface.fetcher_classes:
+ if fetcher_class.can_use_url(url):
+ return fetcher_class.get_pub_light(url)
+
+ # No Module for given url was found
+ raise ValueError("'{}' is not supported".format(url))
+
+ def get_supported_fetchers(self):
+ # print(self.fetcher_classes[0].__name__) Useless right now,
+ # because all classes are called the same
+ return [a.__name__ for a in self.fetcher_classes]
+
+ def import_fetcher_classes(self):
+ """
+ Searches in 'get', if there are [a-z]*.py modules (specific Fetchers)
+ and tries to import them.
+ Saves found modules in 'fetcher_files'.
+ """
+
+ # Path to 'get'-package
+ self.get_path = '{}/get'.format(pathlib.Path(__file__).parent.resolve())
+
+ # Searches for modules with given Pattern
+ fetcher_file_names=[]
+ for file in next(walk(self.get_path), (None, None, []))[2]:
+ if re.match(r'[a-z]+.py', file) is not None:
+ fetcher_file_names.append(file)
+
+ # Tries to import those modules and saves their 'Fetcher'-class
+ for file in fetcher_file_names:
+ try:
+ fetcher_class = importlib.import_module("input.get.{}".format(file[:-3]))
+ try:
+ self.fetcher_classes.append(fetcher_class.__getattribute__('Fetcher'))
+ except Exception as error:
+ ImportError("Module '{}' does not have a 'Fetcher'-class".format(file[:-3]))
+ except Exception:
+ raise ImportError("Module '{}' can not be imported".format(file[:-3]))
diff --git a/input/publication.py b/input/publication.py
new file mode 100755
index 0000000000000000000000000000000000000000..fc512e7173a84695ea566706784c565a7b5ebb8f
--- /dev/null
+++ b/input/publication.py
@@ -0,0 +1,143 @@
+#!/usr/bin/env python3
+
+# this is needed for typing pre python 3.9, this maybe as an large Overhead
+from typing import Any, List
+
+
+class Publication:
+ """
+ Represents a Publications
+ """
+ def __init__(self, doi_url: str, title: str \
+ , contributors: List[str], journal: str \
+ , publication_date: str, subjects: List[str]\
+ , references: List[Any] = None, citations: List[Any] = None ):
+ """
+ Parameters
+ ----------
+ :param doi_url: doi_url of the publication
+ :type doi_url: str
+ :param title: title of the publication
+ :type title: str
+ :param contributors:list of all contributors
+ :type contributors: list[]
+ :param published: date of release
+ :type published: str
+ :param subjects: the subject of the Publication
+ :type subjects: List[str]
+ :param references: the Citation which is been referenced by this Publication
+ :type references: List[Any]
+ :param citations: the Citation which references this Publication
+ :type citations: List[Any]
+ :return: None
+ """
+ self.doi_url = doi_url
+ self.title = title
+ self.contributors = contributors
+ self.journal = journal
+ self.publication_date = publication_date
+ self.subjects = subjects
+ if references is None:
+ self.references = []
+ else:
+ self.references = references
+ if citations is None:
+ self.citations = []
+ else:
+ self.citations = citations
+
+ # For the 'Verarbeitungsgruppe'
+ self.group = None
+
+ def __str__(self) -> str:
+ return ("Title: {}\n"
+ "Doi-url: {}\n"
+ "Authors: {}\n"
+ "Journal: {}\n"
+ "Published on: {}\n"
+ "Subjects: {}\n"
+ "References: \n{}\n"
+ "Citations: \n{}")\
+ .format(self.title, self.doi_url, ", ".join(self.contributors)
+ , self.journal, self.publication_date
+ , ", ".join(self.subjects)
+ , "\n".join(self.get_citation_string(self.references))
+ , "\n".join(self.get_citation_string(self.citations)))
+
+ @staticmethod
+ def get_citation_string(citations):
+ if citations == []:
+ return ["None"]
+ else:
+ citation_string = []
+ for citation in citations:
+ citation_string.append(citation.__str__())
+ return citation_string
+
+ def add_citations(self, citation) -> None:
+ """
+ Appends a list of Citations or Citation to self.citations.
+
+ Parameter
+ ---------
+ :param citation: Citation or Reference of the Publication
+ :type citation: Citation or list[Citation]
+ :return: self.citations
+ """
+ if type(citation) is Citation:
+ self.citations.append(citation)
+
+ # Checks if 'citation' is a list of Citations
+ elif type(citation) is list:
+ for _cit in citation:
+ if type(_cit) is Citation:
+ self.citations.append(_cit)
+ else:
+ raise TypeError("_set_citation expects Citations or List of Citations, not: '{}'"
+ .format(type(_cit)))
+ else:
+ raise TypeError("_set_citation expects Citations or List of Citations, not: '{}'"
+ .format(type(citation)))
+
+ return self.citations
+
+ def __eq__(self, other) -> bool:
+ """ Compares the unique doi_url of two Publications"""
+ if type(self)==type(other):
+ return self.doi_url == other.doi_url
+ return False
+
+
+class Citation:
+ def __init__(self, doi_url: str, title: str \
+ , journal: str, contributors: List[str] \
+ , cit_type: str = "Citation"):
+ """
+ Parameters
+ ----------
+ :param doi_url: doi_url of the publication
+ :type doi_url: str
+ :param title: title of the publication
+ :type title: str
+ :param contributors: list of all contributors
+ :type contributors: List[str]
+ :param cit_type: Specifies if Reference or Citation
+ :type cit_type: str
+ :return: None
+ """
+
+ self.title = title
+ self.doi_url = doi_url
+ self.journal = journal
+ self.contributors = contributors
+ self.cit_type = cit_type
+
+ def __str__(self) -> str:
+ return ("\t{}-Title: {}\n"
+ "\t{}-Doi: {}\n"
+ "\t{}-Journal: {}\n"
+ "\t{}-Contributors: {}\n")\
+ .format(self.cit_type, self.title
+ , self.cit_type, self.doi_url
+ , self.cit_type, self.journal
+ , self.cit_type, ", ".join(self.contributors))
diff --git a/input/requirements.txt b/input/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..a151126691e7f0a9f1c824e9cbac243a96b32e71
--- /dev/null
+++ b/input/requirements.txt
@@ -0,0 +1,2 @@
+beautifulsoup4
+requests
\ No newline at end of file
diff --git a/input/test/__init__.py b/input/test/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/input/test/test_acs.py b/input/test/test_acs.py
new file mode 100644
index 0000000000000000000000000000000000000000..e3dfe84a09d3599de32efbab0dd60655b5414152
--- /dev/null
+++ b/input/test/test_acs.py
@@ -0,0 +1,303 @@
+#!/usr/bin/env python
+
+from input.get.acs import Fetcher as Acs
+from input.publication import Publication, Citation
+from input.test.test_input import FetcherTestCase
+
+
+class AcsTestCase(FetcherTestCase):
+ """
+ Methods with test_* will be detected by unittest and run.
+ """
+
+ def test_acs_url(self):
+ # Positive Testing
+ self.can_use_url_test(Acs, "https://doi.org/10.1021/acs.jcim.1c00203" , True)
+ self.can_use_url_test(Acs, "doi.org/10.1021/acs.jcim.1c00203" , True)
+ self.can_use_url_test(Acs, "10.1021/acs.jcim.1c00203" , True)
+ self.can_use_url_test(Acs, " 10.1021/acs.jcim.1c00203" , True)
+ self.can_use_url_test(Acs, "10.1021/acs.jcim.1c00203 " , True)
+ self.can_use_url_test(Acs, "\t 10.1021/acs.jcim.1c00203 \t\n" , True)
+ self.can_use_url_test(Acs, "https://pubs.acs.org/doi/10.1021/acs.jcim.1c00203" , True)
+
+ # Negative Testing
+ self.can_use_url_test(Acs, "" , False)
+ self.can_use_url_test(Acs, "https://doi.org/10.1038/219021a0" , False)
+ self.can_use_url_test(Acs, "https://www.nature.com/articles/219021a0" , False)
+ self.can_use_url_test(Acs, "https://pubs.acs.org/doi/doi.org/10.1021/acs.jcim.1c00203", False)
+
+
+
+ def test_acs_publication(self):
+ url = "https://doi.org/10.1021/acs.jcim.1c00203"
+ self.get_publication_test(Acs, url, self.expectedPubs[url])
+
+ def test_acs_exceptions(self):
+ test_url= "https://doi.org/10.1021/acs.jcim.1c002"
+ self.get_publication_exception_test(Acs, test_url)
+
+ # Dictionary of Expected Results, with url
+ expectedPubs = {
+ "https://doi.org/10.1021/acs.jcim.1c00203":
+ Publication(
+ doi_url = "https://doi.org/10.1021/acs.jcim.1c00203",
+ title = "AutoDock Vina 1.2.0: New Docking Methods, Expanded Force Field, and Python Bindings",
+ contributors = ["Jerome Eberhardt", "Diogo Santos-Martins", "Andreas F. Tillack", "Stefano Forli"],
+ journal="Journal of Chemical Information and Modeling",
+ publication_date = "July 19, 2021",
+ subjects = ["Algorithms","Ligands","Molecules","Receptors","Macrocycles"],
+ references = [
+ Citation(doi_url = "https://doi.org/10.1002/jcc.21334"
+ , title ="AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading"
+ , journal="Journal of Computational Chemistry"
+ , contributors=["Trott, O.", "Olson, A. J."]
+ , cit_type="Reference")
+ , Citation(doi_url = "https://doi.org/10.1038/nprot.2016.051"
+ , title ="Computational protein-ligand docking and virtual drug screening with the AutoDock suite"
+ , journal="Journal of Natural Products"
+ , contributors=["Forli, S.","Huey, R.","Pique, M. E.","Sanner, M. F.","Goodsell, D. S.","Olson, A. J."]
+ , cit_type="Reference")
+ , Citation(title = "A semiempirical free energy force field with charge-based desolvation"
+ , doi_url = "https://doi.org/10.1002/jcc.20634"
+ , journal="Journal of Computational Chemistry"
+ , contributors=["Huey, R.","Morris, G. M.","Olson, A. J.","Goodsell, D. S."]
+ , cit_type="Reference")
+ , Citation(title="Accelerating autodock4 with gpus and gradient-based local search"
+ , doi_url="https://doi.org/10.1021/acs.jctc.0c01006"
+ , journal="Journal of Chemical Theory and Computation"
+ , contributors=["Santos-Martins, D.","Solis-Vasquez, L.","Tillack, A. F.","Sanner, M. F.","Koch, A.","Forli, S."]
+ , cit_type="Reference")
+ , Citation(title="AutoDockFR: Advances in Protein-Ligand Docking with Explicitly Specified Binding Site Flexibility"
+ , doi_url="https://doi.org/10.1371/journal.pcbi.1004586"
+ , journal="PLoS Computational Biology"
+ , contributors=["Ravindranath, P. A.","Forli, S.","Goodsell, D. S.","Olson, A. J.","Sanner, M. F."]
+ , cit_type="Reference")
+ , Citation(title="Docking flexible cyclic peptides with AutoDock CrankPep"
+ , doi_url="https://doi.org/10.1021/acs.jctc.9b00557"
+ , journal="Journal of Chemical Theory and Computation"
+ , contributors=["Zhang, Y.","Sanner, M. F."]
+ , cit_type="Reference")
+ , Citation(title="Fast, accurate, and reliable molecular docking with QuickVina 2"
+ , doi_url="https://doi.org/10.1093/bioinformatics/btv082"
+ , journal="Bioinformatics"
+ , contributors=["Alhossary, A.","Handoko, S. D.","Mu, Y.","Kwoh, C.-K."]
+ , cit_type="Reference")
+ , Citation(title="Lessons learned in empirical scoring with smina from the CSAR 2011 benchmarking exercise"
+ , doi_url="https://doi.org/10.1021/ci300604z"
+ , journal="Journal of Chemical Information and Modeling"
+ , contributors=["Koes, D. R.","Baumgartner, M. P.","Camacho, C. J."]
+ , cit_type="Reference")
+ , Citation(title="Vina-Carb: Improving Glycosidic Angles during Carbohydrate Docking"
+ , doi_url="https://doi.org/10.1021/acs.jctc.5b00834"
+ , journal="Journal of Chemical Theory and Computation"
+ , contributors=["Nivedha, A. K.","Thieker, D. F.","Makeneni, S.","Hu, H.","Woods, R. J."]
+ , cit_type="Reference")
+ , Citation(title="AutoDock VinaXB: implementation of XBSF, new empirical halogen bond scoring function, into AutoDock Vina"
+ , doi_url="https://doi.org/10.1186/s13321-016-0139-1"
+ , journal="Journal of Cheminformatics"
+ , contributors=["Koebel, M. R.","Schmadeke, G.","Posner, R. G.","Sirimulla, S."]
+ , cit_type="Reference")
+ , Citation(title="Vinardo: A Scoring Function Based on Autodock Vina Improves Scoring, Docking, and Virtual Screening"
+ , doi_url="https://doi.org/10.1371/journal.pone.0155183"
+ , journal="PLoS One"
+ , contributors=["Quiroga, R.","Villarreal, M. A."]
+ , cit_type="Reference")
+ , Citation(title="Lennard-Jones potential and dummy atom settings to overcome the AUTODOCK limitation in treating flexible ring systems"
+ , doi_url="https://doi.org/10.1021/ci700036j"
+ , journal="Journal of Chemical Information and Modeling"
+ , contributors=["Forli, S.","Botta, M."]
+ , cit_type="Reference")
+ , Citation(title="AutoDock4Zn: an improved AutoDock force field for small-molecule docking to zinc metalloproteins"
+ , doi_url="https://doi.org/10.1021/ci500209e"
+ , journal="Journal of Chemical Information and Modeling"
+ , contributors=["Santos-Martins, D.","Forli, S.","Ramos, M. J.","Olson, A. J."]
+ , cit_type="Reference")
+ , Citation(title="A force field with discrete displaceable waters and desolvation entropy for hydrated ligand docking"
+ , doi_url="https://doi.org/10.1021/jm2005145"
+ , journal="Journal of Medicinal Chemistry"
+ , contributors=["Forli, S.","Olson, A. J."]
+ , cit_type="Reference")
+ , Citation(title="Directional phosphorylation and nuclear transport of the splicing factor SRSF1 is regulated by an RNA recognition motif"
+ , doi_url="https://doi.org/10.1016/j.jmb.2016.04.009"
+ , journal="Journal of Molecular Biology"
+ , contributors=["Serrano, P.","Aubol, B. E.","Keshwani, M. M.","Forli, S.","Ma, C.-T.","Dutta, S. K.","Geralt, M.","Wüthrich, K.","Adams, J. A."]
+ , cit_type="Reference")
+ , Citation(title="Covalent docking using autodock: Two-point attractor and flexible side chain methods"
+ , doi_url="https://doi.org/10.1002/pro.2733"
+ , journal="Protein Science"
+ , contributors=["Bianco, G.","Forli, S.","Goodsell, D. S.","Olson, A. J."]
+ , cit_type="Reference")
+ , Citation(title="Consensus docking: improving the reliability of docking in a virtual screening context"
+ , doi_url="https://doi.org/10.1021/ci300399w"
+ , journal="Journal of Chemical Information and Modeling"
+ , contributors=["Houston, D. R.","Walkinshaw, M. D."]
+ , cit_type="Reference")
+ , Citation(title="DockBench: an integrated informatic platform bridging the gap between the robust validation of docking protocols and virtual screening simulations"
+ , doi_url="https://doi.org/10.3390/molecules20069977"
+ , journal="Molecules"
+ , contributors=["Cuzzolin, A.","Sturlese, M.","Malvacio, I.","Ciancetta, A.","Moro, S."]
+ , cit_type="Reference")
+ , Citation(title="A new force field for molecular mechanical simulation of nucleic acids and proteins"
+ , doi_url="https://doi.org/10.1021/ja00315a051"
+ , journal="Journal of the American Chemical Society"
+ , contributors=["Weiner, S. J.","Kollman, P. A.","Case, D. A.","Singh, U. C.","Ghio, C.","Alagona, G.","Profeta, S.","Weiner, P."]
+ , cit_type="Reference")
+ , Citation(title="AutoDock Bias: improving binding mode prediction and virtual screening using known protein-ligand interactions"
+ , doi_url="https://doi.org/10.1093/bioinformatics/btz152"
+ , journal="Bioinformatics"
+ , contributors=["Arcon, J. P.","Modenutti, C. P.","Avendaño, D.","Lopez, E. D.","Defelipe, L. A.","Ambrosio, F. A.","Turjanski, A. G.","Forli, S.","Marti, M. A."]
+ , cit_type="Reference")
+ , Citation(title="Inhomogeneous Fluid Approach to Solvation Thermodynamics. 1. Theory"
+ , doi_url="https://doi.org/10.1021/jp9723574"
+ , journal="Journal of Physical Chemistry B"
+ , contributors=["Lazaridis, T."]
+ , cit_type="Reference")
+ , Citation(title="Inhomogeneous fluid approach to solvation thermodynamics. 2. Applications to simple fluids"
+ , doi_url="https://doi.org/10.1021/jp972358w"
+ , journal="Journal of Physical Chemistry B"
+ , contributors=["Lazaridis, T."]
+ , cit_type="Reference")
+ , Citation(title="Grid inhomogeneous solvation theory: Hydration structure and thermodynamics of the miniature receptor cucurbit[7]uril"
+ , doi_url="https://doi.org/10.1063/1.4733951"
+ , journal="Journal of Chemical Physics"
+ , contributors=["Nguyen, C. N.","Young, T. K.","Gilson, M. K."]
+ , cit_type="Reference")
+ , Citation(title="AutoDock-GIST: Incorporating Thermodynamics of Active-Site Water into Scoring Function for Accurate Protein-Ligand Docking"
+ , doi_url="https://doi.org/10.3390/molecules21111604"
+ , journal="Molecules"
+ , contributors=["Uehara, S.","Tanaka, S."]
+ , cit_type="Reference")
+ , Citation(title="ZINC20—A Free Ultralarge-Scale Chemical Database for Ligand Discovery"
+ , doi_url="https://doi.org/10.1021/acs.jcim.0c00675"
+ , journal="Journal of Chemical Information and Modeling"
+ , contributors=["Irwin, J. J.","Tang, K. G.","Young, J.","Dandarchuluun, C.","Wong, B. R.","Khurelbaatar, M.","Moroz, Y. S.","Mayfield, J.","Sayle, R. A."]
+ , cit_type="Reference")
+ , Citation(title="Structural biology-inspired discovery of novel KRAS–PDEδ inhibitors"
+ , doi_url="https://doi.org/10.1021/acs.jmedchem.7b01243"
+ , journal="Journal of Medicinal Chemistry"
+ , contributors=["Jiang, Y.","Zhuang, C.","Chen, L.","Lu, J.","Dong, G.","Miao, Z.","Zhang, W.","Li, J.","Sheng, C."]
+ , cit_type="Reference")
+ , Citation(title="D3R grand challenge 2015: evaluation of protein–ligand pose and affinity predictions"
+ , doi_url="https://doi.org/10.1007/s10822-016-9946-8"
+ , journal="Journal of Computer-Aided Molecular Design"
+ , contributors=["Gathiaka, S.","Liu, S.","Chiu, M.","Yang, H.","Stuckey, J. A.","Kang, Y. N.","Delproposto, J.","Kubish, G.","Dunbar, J. B.","Carlson, H. A.","Burley, S. K.","Walters, W. P.","Amaro, R. E.","Feher, V. A.","Gilson, M. K."]
+ , cit_type="Reference")
+ , Citation(title="D3R grand challenge 4: blind prediction of protein–ligand poses, affinity rankings, and relative binding free energies"
+ , doi_url="https://doi.org/10.1007/s10822-020-00289-y"
+ , journal="Journal of Computer-Aided Molecular Design"
+ , contributors=["Parks, C. D.","Gaieb, Z.","Chiu, M.","Yang, H.","Shao, C.","Walters, W. P.","Jansen, J. M.","McGaughey, G.","Lewis, R. A.","Bembenek, S. D.","Ameriks, M. K.","Mirzadegan, T.","Burley, S. K.","Amaro, R. E.","Gilson, M. K."]
+ , cit_type="Reference")
+ , Citation(title="D3R Grand Challenge 4: prospective pose prediction of BACE1 ligands with AutoDock-GPU"
+ , doi_url="https://doi.org/10.1007/s10822-019-00241-9"
+ , journal="Journal of Computer-Aided Molecular Design"
+ , contributors=["Santos-Martins, D.","Eberhardt, J.","Bianco, G.","Solis-Vasquez, L.","Ambrosio, F. A.","Koch, A.","Forli, S."]
+ , cit_type="Reference")
+ , Citation(title="Comparison of affinity ranking using AutoDock-GPU and MM-GBSA scores for BACE-1 inhibitors in the D3R Grand Challenge 4"
+ , doi_url="https://doi.org/10.1007/s10822-019-00240-w"
+ , journal="Journal of Computer-Aided Molecular Design"
+ , contributors=["El Khoury, L.","Santos-Martins, D.","Sasmal, S.","Eberhardt, J.","Bianco, G.","Ambrosio, F. A.","Solis-Vasquez, L.","Koch, A.","Forli, S.","Mobley, D. L."]
+ , cit_type="Reference")
+ , Citation(title="Macrocycle modeling in ICM: benchmarking and evaluation in D3R Grand Challenge 4"
+ , doi_url="https://doi.org/10.1007/s10822-019-00225-9"
+ , journal="Journal of Computer-Aided Molecular Design"
+ , contributors=["Lam, P. C.-H.","Abagyan, R.","Totrov, M."]
+ , cit_type="Reference")
+ , Citation(title="Directory of useful decoys, enhanced (DUD-E): better ligands and decoys for better benchmarking"
+ , doi_url="https://doi.org/10.1021/jm300687e"
+ , journal="Journal of Medicinal Chemistry"
+ , contributors=["Mysinger, M. M.","Carchia, M.","Irwin, J. J.","Shoichet, B. K."]
+ , cit_type="Reference")
+ , Citation(title="Evaluation of AutoDock and AutoDock Vina on the CASF-2013 benchmark"
+ , doi_url="https://doi.org/10.1021/acs.jcim.8b00312"
+ , journal="Journal of Chemical Information and Modeling"
+ , contributors=["Gaillard, T."]
+ , cit_type="Reference")
+ , Citation(title="Autodock vina adopts more accurate binding poses but autodock4 forms better binding affinity"
+ , doi_url="https://doi.org/10.1021/acs.jcim.9b00778"
+ , journal="Journal of Chemical Information and Modeling"
+ , contributors=["Nguyen, N. T.","Nguyen, T. H.","Pham, T. N. H.","Huy, N. T.","Bay, M. V.","Pham, M. Q.","Nam, P. C.","Vu, V. V.","Ngo, S. T."]
+ , cit_type="Reference")
+ , Citation(title="Development and validation of a genetic algorithm for flexible docking"
+ , doi_url="https://doi.org/10.1006/jmbi.1996.0897"
+ , journal="Journal of Molecular Biology"
+ , contributors=["Jones, G.","Willett, P.","Glen, R. C.","Leach, A. R.","Taylor, R."]
+ , cit_type="Reference")
+ , Citation(title="Glide: a new approach for rapid, accurate docking and scoring. 1. Method and assessment of docking accuracy"
+ , doi_url="https://doi.org/10.1021/jm0306430"
+ , journal="Journal of Medicinal Chemistry"
+ , contributors=["Friesner, R. A.","Banks, J. L.","Murphy, R. B.","Halgren, T. A.","Klicic, J. J.","Mainz, D. T.","Repasky, M. P.","Knoll, E. H.","Shelley, M.","Perry, J. K."]
+ , cit_type="Reference")
+ , Citation(title="Surflex: fully automatic flexible molecular docking using a molecular similarity-based search engine"
+ , doi_url="https://doi.org/10.1021/jm020406h"
+ , journal="Journal of Medicinal Chemistry"
+ , contributors=["Jain, A. N."]
+ , cit_type="Reference")
+ , Citation(title="A fast flexible docking method using an incremental construction algorithm"
+ , doi_url="https://doi.org/10.1006/jmbi.1996.0477"
+ , journal="Journal of Molecular Biology"
+ , contributors=["Rarey, M.","Kramer, B.","Lengauer, T.","Klebe, G."]
+ , cit_type="Reference")
+ , Citation(title="EDock: blind protein–ligand docking by replica-exchange monte carlo simulation"
+ , doi_url="https://doi.org/10.1186/s13321-020-00440-9"
+ , journal="Journal of Cheminformatics"
+ , contributors=["Zhang, W.","Bell, E. W.","Yin, M.","Zhang, Y."]
+ , cit_type="Reference")
+ , Citation(title="DOCK 6: Impact of new features and current docking performance"
+ , doi_url="https://doi.org/10.1002/jcc.23905"
+ , journal="Journal of Computational Chemistry"
+ , contributors=["Allen, W. J.","Balius, T. E.","Mukherjee, S.","Brozell, S. R.","Moustakas, D. T.","Lang, P. T.","Case, D. A.","Kuntz, I. D.","Rizzo, R. C."]
+ , cit_type="Reference")
+ , Citation(title="Improving scoring-docking-screening powers of protein–ligand scoring functions using random forest"
+ , doi_url="https://doi.org/10.1002/jcc.24667"
+ , journal="Journal of Computational Chemistry"
+ , contributors=["Wang, C.","Zhang, Y."]
+ , cit_type="Reference")
+ , Citation(title="ID-Score: a new empirical scoring function based on a comprehensive set of descriptors related to protein–ligand interactions"
+ , doi_url="https://doi.org/10.1021/ci300493w"
+ , journal="Journal of Chemical Information and Modeling"
+ , contributors=["Li, G.-B.","Yang, L.-L.","Wang, W.-J.","Li, L.-L.","Yang, S.-Y."]
+ , cit_type="Reference")
+ , Citation(title="Further development and validation of empirical scoring functions for structure-based binding affinity prediction"
+ , doi_url="https://doi.org/10.1023/a:1016357811882"
+ , journal="Journal of Computer-Aided Molecular Design"
+ , contributors=["Wang, R.","Lai, L.","Wang, S."]
+ , cit_type="Reference")
+ , Citation(title="A knowledge-based energy function for protein- ligand, protein- protein, and protein- DNA complexes"
+ , doi_url="https://doi.org/10.1021/jm049314d"
+ , journal="Journal of Medicinal Chemistry"
+ , contributors=["Zhang, C.","Liu, S.","Zhu, Q.","Zhou, Y."]
+ , cit_type="Reference")
+ , Citation(title="DLIGAND2: an improved knowledge-based energy function for protein–ligand interactions using the distance-scaled, finite, ideal-gas reference state"
+ , doi_url="https://doi.org/10.1186/s13321-019-0373-4"
+ , journal="Journal of Cheminformatics"
+ , contributors=["Chen, P.","Ke, Y.","Lu, Y.","Du, Y.","Li, J.","Yan, H.","Zhao, H.","Zhou, Y.","Yang, Y."]
+ , cit_type="Reference")
+ , Citation(title="Comparing AutoDock and Vina in ligand/decoy discrimination for virtual screening"
+ , doi_url="https://doi.org/10.3390/app9214538"
+ , journal="Applied Science"
+ , contributors=["Vieira, T. F.","Sousa, S. F."]
+ , cit_type="Reference")
+ , Citation(title="Benchmark of four popular virtual screening programs: construction of the active/decoy dataset remains a major determinant of measured performance"
+ , doi_url="https://doi.org/10.1186/s13321-016-0167-x"
+ , journal="Journal of Cheminformatics"
+ , contributors=["Chaput, L.","Martinez-Sanz, J.","Quiniou, E.","Rigolet, P.","Saettel, N.","Mouawad, L."]
+ , cit_type="Reference")
+ , Citation(title="Array programming with NumPy"
+ , doi_url="https://doi.org/10.1038/s41586-020-2649-2"
+ , journal="Nature"
+ , contributors=["Harris, C. R."]
+ , cit_type="Reference")
+ , Citation(title="Matplotlib: A 2D graphics environment"
+ , doi_url="https://doi.org/10.1109/mcse.2007.55"
+ , journal="Computing in Science & Engineering"
+ , contributors=["Hunter, J. D."]
+ , cit_type="Reference")
+ ], citations = [
+ Citation(doi_url = "https://doi.org/10.1021/acsomega.1c04320"
+ , title ="Novel Anti-Hepatitis B Virus Activity of Euphorbia schimperi and Its Quercetin and Kaempferol Derivatives"
+ , journal="ACS Omega"
+ , contributors=["Mohammad K. Parvez","Sarfaraz Ahmed","Mohammed S. Al-Dosari","Mazin A. S. Abdelwahid","Ahmed H. Arbab","Adnan J. Al-Rehaily","Mai M. Al-Oqail"],cit_type="Citation"),
+
+ ]
+ )
+ }
\ No newline at end of file
diff --git a/input/test/test_input.py b/input/test/test_input.py
new file mode 100755
index 0000000000000000000000000000000000000000..b2ca55f961565fd1192b72ce992c9ff95bd23020
--- /dev/null
+++ b/input/test/test_input.py
@@ -0,0 +1,82 @@
+import unittest
+from input.get.journal_fetcher import JournalFetcher
+from input.interface import InputInterface
+from input.publication import Publication
+
+"""
+Testing the Publication fetcher
+
+Publication 1: 'https://doi.org/10.1021/acs.jcim.1c00203'
+Publication 2: 'doi.org/10.1021/acs.jcim.1c00917'
+Publication 3: '10.1038/nchem.1781'
+Publication 4: '11.12/jaj'
+Publication 5: '11.12/'
+Publication 6: 'https://doi.org/10.1021/acs.jmedchem.0c01332' # Paper is a PDF
+"""
+# TODO: Testcases for:
+# - Specific Journals: Inherit from FetcherTestCase
+# - interface module-importer (test case)
+# - Error detection
+# - wrong/no Journal_fetchers
+# - wrong urls
+# - correct Types in publication
+# - Edgecases (i.e. paper as pdf, no connection, etc)
+
+
+class InterfaceTestCase(unittest.TestCase):
+ def setUp(self):
+ self.assertEqual(InputInterface.instance, None)
+ self.interface = InputInterface()
+
+ def test_singleton(self):
+ # interface should already be made in setUp()
+ self.assertNotEqual(self.interface.instance, None)
+ new_interface = InputInterface()
+ self.assertEqual(self.interface, new_interface)
+
+ # def test_imported_modules(self):
+ # fetchers = self.interface.get_supported_fetchers
+
+class FetcherTestCase(unittest.TestCase):
+
+
+ def can_use_url_test(self, fetcher : JournalFetcher, test_url: str, expected_res: bool):
+ # Tests the 'can_use_url'-method
+ self.assertEqual(fetcher.can_use_url(test_url), expected_res)
+
+
+ def get_publication_test(self, fetcher : JournalFetcher, test_url: str, expected_res: Publication):
+ """
+ this test asserts that every variable is equals to the expected result
+ """
+ actual_res = fetcher.get_publication(test_url)
+ self.assertEqual(actual_res.doi_url, expected_res.doi_url)
+ self.assertEqual(actual_res.title, expected_res.title)
+ self.assertEqual(actual_res.contributors, expected_res.contributors)
+ self.assertEqual(actual_res.journal, expected_res.journal)
+ self.assertEqual(actual_res.publication_date, expected_res.publication_date)
+ self.assertEqual(actual_res.subjects, expected_res.subjects)
+
+ # Checking for all references
+ self.assertEqual(len(actual_res.references), len(expected_res.references))
+ num_references = len(expected_res.references)
+ for i in range(num_references):
+ self.assertEqual(actual_res.references[i].doi_url, expected_res.references[i].doi_url)
+ self.assertEqual(actual_res.references[i].journal, expected_res.references[i].journal)
+ self.assertEqual(actual_res.references[i].contributors, expected_res.references[i].contributors)
+ self.assertEqual(actual_res.references[i].cit_type, expected_res.references[i].cit_type)
+
+ # Checking for all citations
+ self.assertEqual(len(actual_res.citations), len(expected_res.citations))
+ num_citations = len(expected_res.citations)
+ for i in range(num_citations):
+ self.assertEqual(actual_res.citations[i].doi_url, expected_res.citations[i].doi_url)
+ self.assertEqual(actual_res.citations[i].journal, expected_res.citations[i].journal)
+ self.assertEqual(actual_res.citations[i].contributors, expected_res.citations[i].contributors)
+ self.assertEqual(actual_res.citations[i].cit_type, expected_res.citations[i].cit_type)
+
+
+ def get_publication_exception_test(self, fetcher: JournalFetcher, test_url: str):
+ # Ckecks
+ with self.assertRaises(ValueError):
+ fetcher.get_publication(test_url)
\ No newline at end of file
diff --git a/ui_programm_fragmente/input_to_checklist.py b/ui_programm_fragmente/input_to_checklist.py
index af6c0b18c78d007e00b5e042e8cd17cf70137905..3c00ed4c23c7acf914c02af576fec41d8ba2efc7 100644
--- a/ui_programm_fragmente/input_to_checklist.py
+++ b/ui_programm_fragmente/input_to_checklist.py
@@ -4,6 +4,8 @@ from dash import html
from dash import callback_context
from dash.dependencies import Input, Output, State
from dash.exceptions import PreventUpdate
+from input.interface import InputInterface
+import input.publication
app = dash.Dash(__name__)
@@ -22,9 +24,10 @@ app.layout = html.Div([
dcc.Input(id='forward-depth',value='1',type='number',min='1',max='10'),
dcc.Input(id='backward-depth',value='1',type='number',min='1',max='10')
]),
- # Layer 2: For the checklist and Remove-/Start-Buttons
+ # Layer 2: For the checklist, Remove-/Start-Buttons and input-error-message
html.Div([
dcc.Checklist(id='input-checklist',options=[],labelStyle = dict(display='block'),value=[]),
+ html.Div(id='input-err',style={'color':'red'}),
html.Button(id='clear-all-button',children='Clear All'),
html.Button(id='clear-selected-button',children='Clear Selected'),
html.Button(id='start-button',children='Generate Graph')
@@ -50,6 +53,7 @@ input-string is required as Output to clear the input box after each input
Output('input-checklist','options'),
Output('input-checklist','value'),
Output('input-string','value'),
+ Output('input-err','children'),
Input('input-string','value'),
Input('clear-all-button','n_clicks'),
Input('clear-selected-button','n_clicks'),
@@ -71,21 +75,28 @@ def update_input_checklist(input_value,btn1,btn2,all_inputs,selected_inputs):
changed_id = [p['prop_id'] for p in callback_context.triggered][0]
# if clear-all-button was pressed:
if 'clear-all-button' in changed_id:
- return list(),list(),''
+ return list(),list(),'',''
# if clear-selected-button was pressed:
if 'clear-selected-button' in changed_id:
all_inputs = [i for i in all_inputs if i['value'] not in selected_inputs]
- return all_inputs,list(),''
+ return all_inputs,list(),'',''
# when the programm is first started:
if input_value == '':
app.layout['input-checklist'].options.clear()
- return list(),list(),''
+ return list(),list(),'',''
# when a new element is added via dcc.Input
- options = all_inputs
- currValues = [x['value'] for x in options]
- if input_value not in currValues:
- options.append({'label':input_value, 'value':input_value})
- return options,selected_inputs,''
+ if 'input-string' in changed_id:
+ options = all_inputs
+ currValues = [x['value'] for x in options]
+ if input_value not in currValues:
+ try:
+ i = InputInterface()
+ pub = i.get_pub_light(input_value)
+ except Exception as err:
+ return options,selected_inputs,'','{}'.format(err)
+ rep_str = pub.contributors[0] + ',' + pub.journal + ',' + pub.publication_date
+ options.append({'label':rep_str, 'value':input_value})
+ return options,selected_inputs,'',''
'''
This callback shows and hides the (first) help-box
diff --git a/ui_programm_fragmente/upload_to_checklist.py b/ui_programm_fragmente/upload_to_checklist.py
index dba9f698c47250b58ceb006095b763b437cee640..9a094f213901a808ad924f4b1ffa87fb87f2f75d 100644
--- a/ui_programm_fragmente/upload_to_checklist.py
+++ b/ui_programm_fragmente/upload_to_checklist.py
@@ -1,7 +1,9 @@
import dash
from dash import dcc
from dash import html
-from dash.dependencies import Input, Output
+from dash.dependencies import Input, Output, State
+import base64
+import re
app = dash.Dash(__name__)
@@ -15,6 +17,7 @@ app.layout = html.Div([
children=html.Div(
["Drag and drop or click to select a file to upload."]
),
+
style={
"width": "30%",
"height": "60px",
@@ -25,27 +28,51 @@ app.layout = html.Div([
"textAlign": "center",
"margin": "10px",
}),
- dcc.Checklist(id='list-of-inputs',labelStyle = dict(display='block'))
+
]),
+ dcc.Checklist(id='input-checklist',options=list(),labelStyle = dict(display='block'),value=[]),
+
])
@app.callback(
- Output('list-of-inputs','options'),
- Input('upload-data','filenames'),
- Input('upload-data','contents')
+ Output('input-checklist','options'),
+ Input('upload-data','filename'),
+ Input('upload-data','contents'),
+ State('input-checklist','options')
)
-def update_input_list(uploaded_filenames,uploaded_file_contents):
- for line in uploaded_file_contents:
- line = line.rstrip()
- list_of_inputs[line] = line
- try:
- stream = open(uploaded_filenames,'r') # this statement may throw an exception
- except IOError as err: # exception stores error message in err
- # now comes the code reacting on the exception
- sys.stderr.write('{}: {}\n'.format(sys.argv[0], err))
- exit(1)
- return [{'label': i, 'value': i} for i in list_of_inputs], ''
+def update_input_list(uploaded_filenames,uploaded_file_contents,all_inputs):
+ if uploaded_file_contents is not None:
+
+
+ string = uploaded_file_contents
+
+ #cutting the first part of the String away to decode
+ found = base64.b64decode(re.search(',(.+?)$', string).group(1))
+ print(found.decode('utf-8'))
+
+ uploaded_file_contents = found.decode('utf-8')
+
+
+ list_of_inputs = (uploaded_file_contents.split())
+ #das hier sollte es untereinander anzeigen, bekomme ich allerdings nicht auf die Seite...
+ #return (*list_of_inputs, sep="\n")
+
+ options = all_inputs
+ if not options:
+ options = list()
+ CurrValues = [x['value'] for x in options]
+
+
+ # würde auch funktionieren
+ # return (found.decode('utf-8'))
+ for i in list_of_inputs:
+ if i not in CurrValues:
+ options.append({'label':i, 'value':i})
+
+
+ return options
if __name__ == '__main__':
app.run_server(debug=True)
+