[SOLVED] Why is grep only partially working in an org-mode source block?

Issue

This is a follow up question about cleaning a SPARQL-dataset.

The dataset is gained by this cell:

#+name: raw-dataset
#+BEGIN_SRC sparql :url https://query.wikidata.org/sparql :format text/csv :cache yes :exports both
SELECT ?wLabel ?pLabel
WHERE
{
  ?p wdt:P31 wd:Q98270496 .
  ?p wdt:P1416 ?w .
  SERVICE wikibase:label { bd:serviceParam wikibase:language "en" . }
}
ORDER BY ASC(?wLabel) ASC(?pLabel)
LIMIT 10
#+END_SRC


#+RESULTS[7981b64721a5ffc448aa7da773ce07ea8dbaf8ac]: raw-dataset
| wLabel                                        | pLabel       |
|-----------------------------------------------+--------------|
| Q105775472                                    | NFDI4Health  |
| Q1117007                                      | NFDI4Health  |
| Q115254989                                    | NFDI4Objects |
| Q1205424                                      | NFDI4Objects |
| Q17575706                                     | NFDI4Objects |
| Academy of Sciences and Humanities in Hamburg | Text+        |
| Academy of Sciences and Literature Mainz      | NFDI4Culture |
| Academy of Sciences and Literature Mainz      | NFDI4Memory  |
| Academy of Sciences and Literature Mainz      | NFDI4Objects |
| Academy of Sciences and Literature Mainz      | Text+        |

As a second step I take a look at the data using zsh:

#+begin_src sh :var data=raw-dataset :shebang "#!/opt/homebrew/bin/zsh" :exports both
echo ${data}
#+end_src

#+RESULTS:
| Q105775472                                    | NFDI4Health  |
| Q1117007                                      | NFDI4Health  |
| Q115254989                                    | NFDI4Objects |
| Q1205424                                      | NFDI4Objects |
| Q17575706                                     | NFDI4Objects |
| Academy of Sciences and Humanities in Hamburg | Text+        |
| Academy of Sciences and Literature Mainz      | NFDI4Culture |
| Academy of Sciences and Literature Mainz      | NFDI4Memory  |
| Academy of Sciences and Literature Mainz      | NFDI4Objects |
| Academy of Sciences and Literature Mainz      | Text+        |

All fine and I can start with the cleaning part, getting rid of all lines containing Q....:

#+begin_src sh :var data=raw-dataset :exports both :shebang "#!/opt/homebrew/bin/zsh"
echo ${data} | grep -L -E "Q[1-9]"
#+end_src

#+RESULTS:

Somehow using the -L does not work. But without (-L) I get the results as expected from the code:

#+begin_src sh :var data=raw-dataset :shebang "#!/opt/homebrew/bin/zsh" :exports both
echo ${data} | grep -E "Q[1-9]"
#+end_src

#+RESULTS:
| Q105775472 | NFDI4Health  |
| Q1117007   | NFDI4Health  |
| Q115254989 | NFDI4Objects |
| Q1205424   | NFDI4Objects |
| Q17575706  | NFDI4Objects |

Question: Why is -L not working and how can I get rid of the lines starting with Q....?

Solution

To exclude lines beginning with Q and then a digit:

• ... | grep -v '^Q[0-9]'
• ... | grep -Pv '^Q\d'

As mentioned in the comment, -L is a rare commandline option (--files-without-match), for excluding entire files. -v is quite common, for excluding lines matching the pattern or patterns. The v, as mentioned in man grep is from --invert-match

Answered By - stevesliva

Answer Checked By - Cary Denson (WPSolving Admin)