MNT: update update_zenodo.py script (#3042)

This is an attempt to update script `.maint/update_zenodo.py`.

- File `contributors.json` appears to have been replaced by a table in
Markdown file `CONTRIBUTORS.md`.
- Files `former.json` and `FORMER.md` luckily point to the same person.
I haven't changed anything here: still reading from `former.json`.

Note that some contributors have two affiliations, and convey that using
a JSON/Python array inside the Markdown array.
```
| Frederick         | Blaise B.    |  | 0000-0001-5832-5279 | ['McLean Hospital Brain Imaging Center, MA, USA', 'Consolidated Department of Psychiatry, Harvard Medical School, MA, USA']                  |
```
```
| Ghosh             | Satrajit S.  |  | 0000-0002-5312-6729 | ['McGovern Institute for Brain Research, MIT, MA, USA', 'Department of Otolaryngology, Harvard Medical School, MA, USA']                     |
```
Would it be preferable to revert 059b59c use a JSON `contributors.json`
file and have a script update `CONTRIBUTORS.md`?

Alternatively, I can use the “ugly” solution of separating affiliations
in the Markdown table with **`<br/>`**. Probably uglier, I could also
interpret the JSON/Python arrays inside the Markdown table.

.maint/update_zenodo.py

@@ -85,6 +85,36 @@ def get_git_lines(fname='line-contributors.txt'):
     return [' '.join(line.strip().split()[1:-1]) for line in lines if '%' in line]
 
 
+def loads_table_from_markdown(s):
+    """Read the first table from a Markdown text."""
+    table = []
+    header = None
+    for line in s.splitlines():
+        if line.startswith("|"):
+            if not header:
+                # read header and strip bold
+                header = [item.strip("* ") for item in line.split('|')[1:-1]]
+            else:
+                values = [item.strip() for item in line.split('|')[1:-1]]
+                if any(any(c != "-" for c in item) for item in values):
+                    table.append(dict(zip(header, values)))
+        elif header:
+            # we have already seen a table, we're past the end of that table
+            break
+    return table
+
+
+def loads_contributors(s):
+    """Reformat contributors read from the Markdown table."""
+    return [
+        {
+            "affiliation": contributor["Affiliation"],
+            "name": "{}, {}".format(contributor["Lastname"], contributor["Name"]),
+            "orcid": contributor["ORCID"],
+        } for contributor in loads_table_from_markdown(s)
+    ]
+
+
 if __name__ == '__main__':
     data = get_git_lines()
 
@@ -96,7 +126,7 @@ def get_git_lines(fname='line-contributors.txt'):
         creators, data,
         exclude=json.loads(Path('.maint/former.json').read_text()),
         last=CREATORS_LAST)
-    contributors = json.loads(Path('.maint/contributors.json').read_text())
+    contributors = loads_contributors(Path('.maint/CONTRIBUTORS.md').read_text())
     zen_contributors, miss_contributors = sort_contributors(
         contributors, data,
         exclude=json.loads(Path('.maint/former.json').read_text()),