AE corrections by page

[funderburkjim]

From the experience of AE corrections in #318, this dictionary is very dirty. There are many errors still remaining, despite the large number of corrections from #318.

Thus, I think that to get a reasonably clean dictionary, we need to systematically review the words on each page.

To that end, a UI has been developed with the aim of making this process of correction of words on a page as efficient as possible.

The main difference from the other correction UIs is that all the lines of a word entry are editable in a
WYSIWIG way (using the tinyMCE javascript library in the background).

I'd like others to try out the current sample of this UI, and make suggestion comments,
This sample deals just with the words on page 1 of AE.

[funderburkjim]

demolink devanagari

demolink slp1

Note: this was developed on local computer, and then uploaded to Cologne. I think it works on Cologne as locally, but have not checked.

Thanks in advance for any feedback.

[drdhaval2785]

There is a lot of space around. Why not show the scanned page on the same page? This will empower reading of scanned and digitized things side by side.

[gasyoun]

There is a lot of space around.

Wasted space, agree.

This will empower reading of scanned and digitized things side by
side.

Agree, that's similar to https://en.wikisource.org/wiki/Page:Sanskrit_Grammar_by_Whitney_p1.djvu/103 but even more advanced because of the editing in WYSIWYG mode.

And before hand cleaning I would propose to do some regex cleanup.

Abate, v. t. ह्रस् c, लघयति (D.), शम् c.

The c, has no dot, but should always have, like in शम् c., can we check for them and if not, add?

What about extracting all devanagari words and comparing to MW, never done before, right? Should we not work with this good looking UI after the big dirt is out? I think it will give more fruit, if we use the old batch methods to weed out the hundreds of spelling mistakes out there.

[funderburkjim]

about extracting all devanagari words and comparing to MW.

This WAS done (and in fact comparisons made to all dictionary headwords) in the prior step of correction (see #318). That step got around 1500 corrections, as I recall.

[funderburkjim]

c, -> c.

Good observation. This is always 'causal'. Will put it on todo list.

[funderburkjim]

The wasted space is not a big deal here.

Since we are dealing with just one scan page at a time, the user can click to open that page in a separate window, do the scan enlargement, and be set for all the cases in the batch.

[funderburkjim]

Sampada's begun working on the pages, and I've done a few.

It seems to take about 30-45 min per page. On the first 9 pages, there were 67 corrections -- that would
be 7 per page.

If @juhnowski @SergeA or others preferring Devanagari want to do any, please let me know, so work can be
coordinated with what Sampada is doing. I'm planning to do installation about every 10 pages, so not all pages are prepared at once.

The UI should already support Devanagari, but only limited testing has been done with Devanagari. Although the correction UI is WYSIWIG, there are a couple of
quirks of data entry that a corrector needs to understand so that corrections are properly interpreted.

[gasyoun]

It seems to take about 30-45 min per page. On the first 9 pages, there were 67 corrections -- that would
be 7 per page.

That's a lot both ways. Thanks to Sampada.

I'm planning to do installation about every 10 page

Understood.

[funderburkjim]

Meant to say @SergeA in comment above.

[funderburkjim]

Can anyone recognize this word, under headword 'arch':

[gasyoun]

चनुर

[drdhaval2785]

catura with a print smudge.

[funderburkjim]

@drdhaval2785 Thanks. That fits with vidagDa.

[sanskritisampada]

Did not expect to find so many errors !!! 7 per page is too many! Seems to be a good decision to do the detailed checking.

…

On Tue, Mar 28, 2017 at 5:48 AM, Mārcis Gasūns ***@***.***> wrote: It seems to take about 30-45 min per page. On the first 9 pages, there were 67 corrections -- that would be 7 per page. That's a lot both ways. Thanks to Sampada. I'm planning to do installation about every 10 page Understood. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#340 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AKak33yNnlKrKtjfyroaqTNa6Oe02SjQks5rqIMMgaJpZM4MaUuv> .

-- *"Faith,more faith!Faith in your possibilities,faith in the Power that is at work behind the veil,and the offered guidance." - Sri Aurobindo*

[gasyoun]

Did not expect to find so many errors

I did. I only hope that it's the best UI possible, so you spend as little amount of time as possible in doing so. Thanks, Sampada!

[funderburkjim]

In AE, the author typically writes a root as 'masj' (to sink, to bathe) and gives forms as 'majjati', etc.
Also 'masja' in SKD, VCP.

In MW, PW , WIL the root shows as 'majj' .

So far, I haven't found an author who mentions both forms.

Two questions:

• Are we right in thinking 'masj' and 'majj' are the same root?
• Is there some explanation for this difference in different dictionaries?

[gasyoun]

Is there some explanation for this difference in different dictionaries?

In Zaliznyak, the authority above all in Russian on dhatus, it's called majj. And it was just yesterday I was exploring it. But I've never seen 'masj' and 'majj' in one source, it's allomorphs of the same root. I've never seen an explanation. MW, PW are on the right track.

[drdhaval2785]

`wumasjo` is the root in dhAtupATha. `wu` and `o` are anubandhas and are elided. `masj` is what remains. Converted to `majj` from rule स्तोः श्चुना श्चुः.

[gasyoun]

masj is what remains.

So it's not a final product, but a mid-stage byproduct, we could say so, Dhaval?

[funderburkjim]

Really good explanation as to why both forms are acceptable as 'the' root.

[funderburkjim]

question on rakzaRasAna

Under headword bulwark in AE, we see

bulwark [p= 044] : Bulwark, s. वप्रः-प्रं, प्राकारः. 2 आश्रयः, श- 
-रणं, संश्रयस्थानं, आलंबः; रक्षणसानं. [L=1285]

Is rakzaRasAna possible (if so how to understand 'sAna') -- the print shows this.

Or, should it be rakzaRasTAna which makes sense (a bulwark is a 'place of protection' ) ?

[funderburkjim]

question on jrim

Under headword COOL, with sense 'to cool down', AE appears to have 'jrim', maybe. Such a Sanskrit
word is found nowhere.

Could Apte have meant 'jfmB' (MW spelling), one of whose meanings (with causal) is 'cause to feel at ease'?

[funderburkjim]

missing headword 'Invincible'

Sampada discovered a missing headword in the digitization for page 239 of AE.

2388 old
32388 new <P>{@Invincible,@} {%a.%} {#ajawya, durjaya, durAsada, a-#}
32388x ins {#-Dfzya, adamya.#}

To add this without changing subsequent L-numbers is a problem at the moment. Do this when the meta-line form of AE has been accomplished.

This would be best done AFTER Sampada has finished the rest of
page-by-page corrections of AE, probably sometime toward the end of this year.

[gasyoun]

To add this without changing subsequent L-numbers is a problem at the moment

Let's change. We do not care yet for others as MW.

[funderburkjim]

Question re kOwwinya

In AE, under headword 'pander' we find spelling kOwwinyaM:

This spelling is also shown in another edition of AE.

However, only kOwwanya is found in any dictionary (MD,MW,PW,AP).

Should the 'inyam' spelling be considered a print error in AE ?

[SergeA]

कुट्टनी=कुट्टिनी >> कौट्टन्यम्=कौट्टिन्यम्
The spelling कौट्टिन्यं is possible, so there is no problem with errors.
If we'll compare the dics we'll find that there is only one source reference to Rajatarangini given by PW, copied by MW and recopied by AP. So we have 2 possible forms with 1 source for the first and no sources for the other. But it does not mean the second is wrong.

[SergeA]

Concerning the previous question for supposed ज्रिम्

with sense 'to cool down', AE appears to have 'jrim'

another edition by the last link gives perfect reading विरम्

[SergeA]

question on rakzaRasAna

also by the link - रक्षणसाधनं

[funderburkjim]

@SergeA Thanks for explanation and research on these three.

kOwwinyam

My understanding now is that kOwwinyam is derived from kuwwinI by some 'ya' taddhita suffix formation rule which introduces the vriddhi O of u; and kOwwanyam is similarly derived from kuwwanI.

jrim and rakzaRasAna

Have changed to viram and rakzaRasADanaM, per the newer print edition, classifying as print error in print edition used for Cologne digitization.

[gasyoun]

changed to viram and rakzaRasADanaM, per the newer print edition, classifying as print error in print edition used for Cologne digitization.

Long live the Jim.

[funderburkjim]

DONE!

Sampada informed me today that all 501 pages of AE have now been examined and corrections provided.

All in all, there are about 7400 corrections as fruit of this year-long endeavor, approximately 15 corrections per page.

Let's all doff our hat to Sampada for her persistence in seeing this project through. The Cologne
digitization of Apte's English-Sanskrit Dictionary is now immensely better and more useful.

Thanks, Sampada!

[drdhaval2785]

I extend my gratitude to Sampada.

[gasyoun]

7400 corrections - what patience did it take.
Nothing of this spirit is seen in Pune. It would
take 10 Sampadas to finish that dictionary, but
we have only 1, so I bow to her lotus feet.

[funderburkjim]

invincible added

The correction to add 'invincible' as headword (see comment above) has now been made to the meta-line
version of ae. It is L=5754.1. Love those decimal L-numbers :)

[sanskritisampada]

Thank you so much ! Your praise is generous! Everybody in this group is doing their bit to contribute to important foundational research in the field of Sanskrit ... this is my bit! :-) Regards Sampada

…

On Tue, Mar 13, 2018 at 10:12 AM Mārcis Gasūns ***@***.***> wrote: 7400 corrections - what patience did it take. Nothing of this spirit is seen in Pune. It would take 10 Sampadas to finish that dictionary, but we have only 1, so I bow to her lotus feet. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#340 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AKak3zNj8snJTUMokEDe_F7iochD9hZvks5td4zcgaJpZM4MaUuv> .

-- *"Faith,more faith!Faith in your possibilities,faith in the Power that is at work behind the veil,and the offered guidance." - Sri Aurobindo*

[gasyoun]

@sanskritisampada do not try for this elephant to look like a puppy, it's huge! What's next?