add ntv-pandas to the ecosystem (#55421)

* Create 0007-compact-and-reversible-JSON-interface.md

* change PDEP number (7 -> 12)

* Add FAQ to the PDEPS 0012

* Update 0012-compact-and-reversible-JSON-interface.md

* Update 0012-compact-and-reversible-JSON-interface.md

* Update 0012-compact-and-reversible-JSON-interface.md

* pre-commit codespell

* Update 0012-compact-and-reversible-JSON-interface.md

* Update 0012-compact-and-reversible-JSON-interface.md

* delete summary

* delete mermaid flowchart

* with summary, without mermaid flowchart

* rename Annexe -> Appendix

* add tableschema specification

* add orient="table"

* Add Table Schema extension

* Update 0012-compact-and-reversible-JSON-interface.md

* Update 0012-compact-and-reversible-JSON-interface.md

* Update 0012-compact-and-reversible-JSON-interface.md

* Update 0012-compact-and-reversible-JSON-interface.md

* Update 0012-compact-and-reversible-JSON-interface.md

* Update 0012-compact-and-reversible-JSON-interface.md

* add 'ntv-pandas' in the 'ecosystem' file

* Update ecosystem.md

* Update ecosystem.md

* Update ecosystem.md

Author: loco-philippe

web/pandas/community/ecosystem.md

@@ -345,6 +345,27 @@ which pandas excels.
 
 ## IO
 
+### [NTV-pandas](https://github.com/loco-philippe/ntv-pandas)*: A semantic, compact and reversible JSON-pandas converter*
+
+NTV-pandas provides a JSON converter with more data types than the ones supported by pandas directly.
+
+It supports the following data types:
+- pandas data types
+- data types defined in the [NTV format](https://loco-philippe.github.io/ES/JSON%20semantic%20format%20(JSON-NTV).htm)
+- data types defined in [Table Schema specification](http://dataprotocols.org/json-table-schema/#field-types-and-formats)
+
+The interface is always reversible (conversion round trip) with two formats (JSON-NTV and JSON-TableSchema).
+
+*example*
+```python
+import ntv_pandas as npd
+
+jsn = df.npd.to_json(table=False) # save df as a JSON-value (format Table Schema if table is True else format NTV )
+df  = npd.read_json(jsn)  # load a JSON-value as a DataFrame
+
+df.equals(npd.read_json(df.npd.to_json(df))) # True in any case, whether table=True or not
+```
+
 ### [BCPandas](https://github.com/yehoshuadimarsky/bcpandas)
 
 BCPandas provides high performance writes from pandas to Microsoft SQL Server,