gismap 0.2.1__py3-none-any.whl → 0.3.0__py3-none-any.whl

gismap/gisgraphs/style.py

@@ -0,0 +1,119 @@
+from string import Template
+
+
+# language=css
+default_style = Template("""
+
+                         #box-$uid {
+                             position: relative;
+                             width: 100% !important;
+                             height: 80vh !important;
+                             max-width: 100vw !important;
+                             max-height: 80vh !important;
+                             min-height: 80vh;
+                             color: #111;
+                         }
+
+                         #vis-$uid {
+                             height: 100%; /* Make the inner div fill the parent */
+                             width: 100%; /* Make the inner div fill the parent */
+                             box-sizing: border-box;
+                             border: 1px solid #444;
+                         }
+
+                         html[data-theme="dark"] #vis-$uid {
+                             background-color: var(--pst-color-background, #14181e);
+                         }
+
+                         .modal {
+                             display: none;
+                             position: fixed;
+                             z-index: 1000;
+                             left: 0;
+                             top: 0;
+                             width: 100%;
+                             height: 100%;
+                             overflow: auto;
+                             background-color: rgba(10, 10, 10, 0.85);
+                         }
+
+                         .modal-content {
+                             background-color: #f4f4f7;
+                             color: #222235;
+                             margin: 10% auto;
+                             padding: 24px;
+                             border: 1px solid #888;
+                             width: 50%;
+                             border-radius: 8px;
+                             box-shadow: 0 5px 15px rgba(0, 0, 0, .6);
+                         }
+                         .modal a {color: #2958d7;}
+                         .modal a:visited {color: #8435a8;}
+
+                         .close {
+                             color: #777;
+                             float: right;
+                             font-size: 28px;
+                             font-weight: bold;
+                             cursor: pointer;
+                         }
+
+                         .close:hover, .close:focus {
+                             color: #aaa;
+                             text-decoration: none;
+                             cursor: pointer;
+                         }
+
+                         .watermark {
+                             position: absolute;
+                             text-decoration: none;
+                             color: #888;
+                             font-size: min(2vw, 10px);
+                             z-index: 10;
+                         }
+
+                         .gislink {
+                             left: 10px;
+                             bottom: 10px;
+                             pointer-events: auto;
+                         }
+
+                         .button {
+                             background: none;
+                             border: none;
+                             padding: 0;
+                             margin: 0;
+                             cursor: pointer;
+                         }
+
+                         .redraw {
+                             left: 10px;
+                             top: 10px;
+                         }
+
+                         .fullscreen {
+                             bottom: 10px;
+                             right: 10px;
+                         }
+
+                         .legend {
+                             display: inline-block;
+                             padding: 10px 16px;
+                             border-radius: 8px;
+                             box-shadow: 0 2px 8px 0 rgba(0, 0, 0, 0.10);
+                             border: 1px solid var(--legend-border, #bbb);
+                             background: var(--jp-layout-color1, #f5f5fa);
+                             background-color: var(--legend-bg, rgba(240, 240, 245, 0.95));
+                             position: absolute;
+                             top: 12px;
+                             right: 12px;
+                             z-index: 20;
+                         }
+
+                         .legend-entry {
+                             display: flex;
+                             margin-right: 10px;
+                             align-items: center;
+                             cursor: pointer;
+                         }
+                         """)

gismap/gisgraphs/widget.py

@@ -0,0 +1,145 @@
+import re
+import unicodedata
+import base64
+from IPython.display import display, HTML
+import ipywidgets as widgets
+from contextlib import contextmanager
+
+from gismap.lab.egomap import EgoMap
+from gismap.lab.labmap import ListMap
+
+
+@contextmanager
+def dummy_context():
+    yield
+
+
+def safe_filename(name):
+    """
+    Parameters
+    ----------
+    name: :class:`str`
+        Pretty much anything.
+
+    Returns
+    -------
+    :class:`str`
+        GisMap filename.
+    """
+    normalized = unicodedata.normalize("NFKD", name)
+    ascii_only = normalized.encode("ascii", "ignore").decode("ascii")
+    ascii_only = ascii_only.replace(" ", "_")
+    safe_str = re.sub(r"[^a-zA-Z0-9_]", "", ascii_only)
+    return f"gismap-{safe_str[:60]}.html"
+
+
+place_holder = "Diego Perino, The-Dang Huynh, François Durand (hal: fradurand, dblp: 38/11269), Rim Kaddah, Leonardo Linguaglossa, Céline Comte"
+
+
+class GismapWidget:
+    """
+    A simple widget to test the production of LabMaps and EgoMaps.
+
+    Examples
+    --------
+
+    This is a doctest example. Use a notebook to play with the widget.
+
+    >>> gw = GismapWidget()  # doctest: +ELLIPSIS +NORMALIZE_WHITESPACE
+    VBox(children=(HTML(value=''), Output(), HBox(children=(Textarea(value='', ...
+    >>> gw.names.value = "Fabien Mathieu"
+    >>> gw.dbs.value = "HAL"
+    >>> gw.size.value = 3
+    >>> gw.compute_function(gw.compute, show=False)
+    >>> gw.save_link.value[:30]
+    "<a href='data:text/html;base64"
+    >>> gw.names.value = "Diego Perino, Laurent Viennot"
+    >>> gw.compute_function(gw.compute, show=False)
+    >>> gw.save_link.value[:30]
+    "<a href='data:text/html;base64"
+    """
+
+    def __init__(self):
+        self.names = widgets.Textarea(
+            placeholder=place_holder,
+            description="Name(s):",
+            layout=widgets.Layout(width="50%", height="100px"),
+        )
+        self.dbs = widgets.RadioButtons(
+            options=["HAL", "DBLP", "Both"],
+            description="DB(s):",
+            layout=widgets.Layout(width="80px", max_width="20%"),
+        )
+        self.size = widgets.IntSlider(
+            value=10,
+            min=1,
+            max=150,
+            step=1,
+            description="Size",
+            layout=widgets.Layout(width="250px"),
+        )
+        self.compute = widgets.Button(
+            description="Map!", layout=widgets.Layout(width="120px", max_width="140px")
+        )
+        self._col = widgets.VBox(
+            [self.size, self.compute],
+            layout=widgets.Layout(
+                align_items="center", max_width="27%", overflow="hidden"
+            ),
+        )
+        self.save_link = widgets.HTML(value="")
+        self.compute.on_click(self.compute_function)
+        self.out = widgets.Output()
+        self.widget = widgets.VBox(
+            [self.save_link, self.out, widgets.HBox([self.names, self._col, self.dbs])]
+        )
+        display(self.widget)
+        self.show = True
+
+    def html(self):
+        dbs = (
+            "hal"
+            if self.dbs.value == "HAL"
+            else "dblp"
+            if self.dbs.value == "DBLP"
+            else ["hal", "dblp"]
+        )
+        name = self.names.value
+        pattern = r",\s*(?![^()]*\))"
+        names = [n.strip() for n in re.split(pattern, name)]
+        self.save_link.value = ""
+        ctx = self.out if self.show else dummy_context()
+        if len(names) > 1:
+            lab = ListMap(names, dbs=dbs, name="planet")
+            if self.show:
+                self.out.clear_output()
+            with ctx:
+                lab.update_authors()
+                lab.update_publis()
+                extra = self.size.value - len(lab.authors)
+                if extra > 0:
+                    lab.expand(target=extra)
+        else:
+            lab = EgoMap(names[0], dbs=dbs)
+            if self.show:
+                self.out.clear_output()
+            with ctx:
+                lab.build(target=self.size.value)
+        return lab.html()
+
+    def compute_function(self, b, show=True):
+        self.show = show
+        full = self.html()
+        b64 = base64.b64encode(
+            f"<html><body>{full}</body></html>".encode("utf8")
+        ).decode("utf8")
+        payload = f"data:text/html;base64,{b64}"
+        savename = safe_filename(self.names.value)
+        link_html = (
+            f"<a href='{payload}' download='{savename}'>Download the Map!</a>"
+        )
+        self.save_link.value = link_html
+        if show:
+            self.out.clear_output()
+            with self.out:
+                display(HTML(full))

gismap/gismo.py

@@ -11,7 +11,7 @@ def make_post_publi(lab):
 
     Parameters
     ----------
-    lab: :class:`~gismap.lab.lab.Lab`
+    lab: :class:`~gismap.lab.labmap.LabMap`
         Lab that contains the corpus publications.
 
     Returns
@@ -32,7 +32,7 @@ def make_gismo(lab, vectorizer_parameters=None):
 
     Parameters
     ----------
-    lab: :class:`~gismap.lab.lab.Lab`
+    lab: :class:`~gismap.lab.labmap.LabMap`
         Lab that contains publications.
     vectorizer_parameters: :class:`dict`
         Overriding parameters for the Countvectorizer of the gismo.

gismap/lab/__init__.py

@@ -1,10 +1,6 @@
-from gismap.lab.lab import (
-    Lab as Lab,
-    ListLab as ListLab,
+from gismap.lab.labmap import (
+    LabMap as Map,
+    ListMap as ListMap,
 )
-from gismap.lab.lip6 import Lip6 as Lip6, Lip6Lab as Lip6Lab
-from gismap.lab.toulouse import Solace as Solace, LaasLab as LaasLab
-from gismap.lab.graph import lab2graph as lab2graph
-from gismap.lab.vis import generate_html as generate_html
 from gismap.lab.egomap import EgoMap as EgoMap
 from gismap.lab.lab_author import LabAuthor as LabAuthor

gismap/lab/egomap.py

@@ -1,12 +1,12 @@
-from gismap.lab.lab import Lab
+from gismap.lab.labmap import LabMap
 from gismap.lab.lab_author import LabAuthor
 
 
-class EgoMap(Lab):
+class EgoMap(LabMap):
     """
     Parameters
     ----------
-    sun
+    star
     args
     kwargs
 
@@ -15,28 +15,27 @@ class EgoMap(Lab):
 
     >>> dang = EgoMap("The-Dang Huynh", dbs="hal")
     >>> dang.build(target=10)
-    >>> sorted(a.name for a in dang.authors.values())  # doctest: +NORMALIZE_WHITESPACE
-    ['Bruno Kauffmann', 'Chung Shue Chen', 'Fabien Mathieu', 'François Baccelli', 'Laurent Viennot', 'Ludovic Noirie',
-    'Siu-Wai Ho', 'Sébastien Tixeuil', 'The-Dang Huynh', 'Yannick Carlinet']
+    >>> sorted(a.name for a in dang.authors.values())  # doctest: +NORMALIZE_WHITESPACE +ELLIPSIS
+    ['Bruno Kauffmann', 'Chung Shue Chen', 'Fabien Mathieu',...
     """
 
-    def __init__(self, sun, *args, **kwargs):
-        if isinstance(sun, str):
-            sun = LabAuthor(sun)
-        sun.metadata.position = (0, 0)
-        self.sun = sun
+    def __init__(self, star, *args, **kwargs):
+        if isinstance(star, str):
+            star = LabAuthor(star)
+        star.metadata.position = (0, 0)
+        star.metadata.group = "star"
+        self.star = star
         super().__init__(*args, **kwargs)
 
     def _author_iterator(self):
-        yield self.sun
+        yield self.star
 
     def build(self, **kwargs):
         target = kwargs.pop("target", 50)
-        group = kwargs.pop("group", "moon")
-        self.update_authors(desc="Sun metadata")
-        self.update_publis(desc="Sun publications")
+        self.update_authors(desc="Star metadata")
+        self.update_publis(desc="Star publications")
+        kwargs["target"] = target - len(self.authors)
+        self.expand(group="planet", desc="Planets", **kwargs)
         kwargs["target"] = target - len(self.authors)
-        self.expand(group=None, desc="Planets", **kwargs)
-        kwargs.update({"target": target - len(self.authors), "group": group})
         if kwargs["target"] > 0:
-            self.expand(desc="Moons", **kwargs)
+            self.expand(group="moon", desc="Moons", **kwargs)

gismap/lab/expansion.py

@@ -53,7 +53,7 @@ def count_prospect_entries(lab):
 
     Parameters
     ----------
-    lab: :class:`~gismap.lab.lab.Lab`
+    lab: :class:`~gismap.lab.labmap.LabMap`
         Reference lab.
 
     Returns
@@ -73,8 +73,8 @@ def count_prospect_entries(lab):
                     count_publications.append(a.key)
                 else:
                     lab_authors.add(a.key)
-            for l in lab_authors:
-                count_coauthors[l].update(new_authors)
+            for a in lab_authors:
+                count_coauthors[a].update(new_authors)
 
     count_coauthors = Counter(
         k for new_authors in count_coauthors.values() for k in new_authors
@@ -117,7 +117,7 @@ def get_prospects(lab):
     """
     Parameters
     ----------
-    lab: :class:`~gismap.lab.lab.Lab`
+    lab: :class:`~gismap.lab.labmap.LabMap`
         Reference lab.
 
     Returns
@@ -150,7 +150,7 @@ def get_member_names(lab):
     """
     Parameters
     ----------
-    lab: :class:`~gismap.lab.lab.Lab`
+    lab: :class:`~gismap.lab.labmap.LabMap`
         Reference lab.
 
     Returns
@@ -222,12 +222,13 @@ def proper_prospects(
         locs = [j for j in np.where(jc[i, :] > threshold)[0] if not done[j]]
         done[locs] = True
         sources = sort_author_sources([prospects[j].author for j in locs])
-        strength = sum(prospects[j].score for j in locs)
-        new_author = LabAuthor.from_sources(sources)
-        new_lab.append((strength, new_author))
+        if sources:
+            strength = sum(prospects[j].score for j in locs)
+            new_author = LabAuthor.from_sources(sources)
+            new_lab.append((strength, new_author))
 
     # Extract top prospects
-    new_lab = [l[1] for l in sorted(new_lab, key=lambda l: l[0], reverse=True)][
+    new_lab = [a[1] for a in sorted(new_lab, key=lambda a: a[0], reverse=True)][
         :max_new
     ]
     new_rosetta = {s.key: a for a in new_lab for s in a.sources}

gismap/lab/filters.py

@@ -1,18 +1,93 @@
-def taboos(txt, words):
+import re
+
+# editorials = re.compile(r"ditorial|foreword", re.IGNORECASE)
+# charlatans = re.compile(r"Raoult|Kofman|Buob")
+
+editorials = ["ditorial", "Foreword", "Brief Announcement", "Preface", "Préface"]
+charlatans = ["Raoult", "Kofman", "Buob"]
+
+
+def re_filter(words):
+    """
+    Parameters
+    ----------
+    words: :class:`list` or :class:`str`
+        List of word(s) to filter.
+
+    Returns
+    -------
+    callable
+        Filter function.
+    """
     if isinstance(words, str):
-        return words not in txt
+        taboo = re.compile(words)
     else:
-        return all(w not in txt for w in words)
+        taboo = re.compile("|".join(words))
+    return lambda txt: taboo.search(txt) is None
+
+
+def publication_size_filter(n_max=9):
+    """
+    Parameters
+    ----------
+    n_max: int, default=9
+        Maximum number of co-authors allowed.
+
+    Returns
+    -------
+    callable
+        Filter on number of co-authors.
+    """
+    return lambda p: len(p.authors) <= n_max
+
+
+def publication_oneword_filter(n_min=2):
+    """
+
+    Parameters
+    ----------
+    n_min: int, default=2
+        Minimum number of words required in the title.
+
+    Returns
+    -------
+    callable
+        Filter on number of words required in the title.
+    """
+    return lambda p: len(p.title.split()) >= n_min
 
-def publication_size_filter(n_max=10):
-    return lambda p: len(p.authors) < n_max
 
 def publication_taboo_filter(w=None):
+    """
+    Parameters
+    ----------
+    w: :class:`list`, optional
+        List of words to filter.
+
+    Returns
+    -------
+    Callable
+        Filter function on publications.
+    """
     if w is None:
-        w = ["Editorial", "Foreword"]
-    return lambda p: taboos(p.title, w)
+        w = editorials
+    regex = re_filter(w)
+    return lambda p: regex(p.title)
+
 
 def author_taboo_filter(w=None):
+    """
+    Parameters
+    ----------
+    w: :class:`list`, optional
+        List of words to filter.
+
+    Returns
+    -------
+    Callable
+        Filter function on authors.
+    """
     if w is None:
-        w = ["Buob", "Kofman"]
-    return lambda a: taboos(a.name, w)
+        w = charlatans
+    regex = re_filter(w)
+    return lambda p: regex(p.name)

gismap/lab/lab_author.py

@@ -1,7 +1,8 @@
 from dataclasses import dataclass, field
+import re
 
 from gismap import get_classes, HAL, DBLP
-from gismap.sources.models import DB
+from gismap.sources.models import DB, db_class_to_auth_class
 from gismap.sources.multi import SourcedAuthor, sort_author_sources
 from gismap.utils.common import LazyRepr, list_of_objects
 from gismap.utils.logger import logger
@@ -26,6 +27,8 @@ class AuthorMetadata(LazyRepr):
         Group of the author.
     position: :class:`tuple`
         Coordinates of the author.
+    keys: :class:`dict`
+        Some DB key values of the author.
     """
 
     url: str = None
@@ -36,6 +39,26 @@ class AuthorMetadata(LazyRepr):
 
 @dataclass(repr=False)
 class LabAuthor(SourcedAuthor):
+    """
+    Examples
+    ---------
+    The metadata and DB key(s) of an author can be entered in parentheses using key/values.
+
+    Improper key/values are ignored (with a warning).
+
+
+    >>> dummy= LabAuthor("My Name(img: https://my.url.img, group:me,url:https://mysite.org,hal:key1,dblp:toto,badkey:hello,no_colon_separator)")
+    >>> dummy.metadata
+    AuthorMetadata(url='https://mysite.org', img='https://my.url.img', group='me')
+    >>> dummy.sources
+    [HALAuthor(name='My Name', key='key1'), DBLPAuthor(name='My Name', key='toto')]
+
+    You can enter multiple keys for the same DB. HAL key types are automatically detected.
+
+    >>> dummy2= LabAuthor("My Name (hal:key1,hal:123456,hal: My Other Name )")
+    >>> dummy2.sources
+    [HALAuthor(name='My Name', key='key1'), HALAuthor(name='My Name', key='123456', key_type='pid'), HALAuthor(name='My Name', key='My Other Name', key_type='fullname')]
+    """
     metadata: AuthorMetadata = field(default_factory=AuthorMetadata)
 
     def auto_img(self):
@@ -45,6 +68,30 @@ class LabAuthor(SourcedAuthor):
                 self.metadata.img = img
                 break
 
+    def __post_init__(self):
+        pattern = r"\s*([^,(]+)\s*(?:\(([^)]*)\))?\s*$"
+        match = re.match(pattern, self.name)
+        if match:
+            self.name = match.group(1).strip()
+            content = match.group(2)
+            if content:
+                for kv in content.split(","):
+                    if ":" not in kv:
+                        logger.warning(f"I don't know what to do with {kv}.")
+                        continue
+                    k, v = kv.split(":", 1)
+                    k = k.strip().lower()
+                    v = v.strip()
+                    if k in db_dict:
+                        DBAuthor = db_class_to_auth_class(db_dict[k])
+                        self.sources.append(DBAuthor(name=self.name, key=v))
+                    elif k in ["url", "img", "group"]:
+                        setattr(self.metadata, k, v)
+                    else:
+                        logger.warning(f"I don't know what to do with {kv}.")
+        else:
+            self.name = self.name.strip()
+
     def auto_sources(self, dbs=None):
         """
         Automatically populate the sources based on author's name.
@@ -63,9 +110,9 @@ class LabAuthor(SourcedAuthor):
         for db in dbs:
             source = db.search_author(self.name)
             if len(source) == 0:
-                logger.warning(f"{self.name} not found in {db.db_name}")
+                logger.info(f"{self.name} not found in {db.db_name}")
             elif len(source) > 1:
-                logger.warning(f"Multiple entries for {self.name} in {db.db_name}")
+                logger.info(f"Multiple entries for {self.name} in {db.db_name}")
             sources += source
         if len(sources) > 0:
             self.sources = sort_author_sources(sources)
@@ -74,7 +121,7 @@ class LabAuthor(SourcedAuthor):
 def labify_author(author, rosetta):
     if isinstance(author, LabAuthor):
         return author
-    return rosetta.get(author.key, author)
+    return rosetta.get(author.key, rosetta.get(author.name, author))
 
 
 def labify_publications(pubs, rosetta):