Links to MBH Calcutta edition

[funderburkjim]

The aim here is to prepare links to pdf pages for the MBH literary source references in pwg dictionary.

We hope to make use of material developed by @Andhrabharati as described here.
The material includes

• 4 pdfs, with about 3000 pages.
• An index containing, for each page, the parvan, and range of verses on the page.

From these, for a given reference in pwg, we should be able to provide a link to a pdf of the MBH page.

[funderburkjim]

reality check.

The first MBH reference in pwg is <ls>MBH. 1, 2523.</ls> under headword aMSa.
The relevant line of index file is 1 92 I 2513 2541 29
which says in volume 1, page 92 contains parvan 'I' (= 1) verses from 2513 to 2541 (29 verses).

And here is a snip of the relevant page:

And indeed the word aMSa occurs where expected. Hurray!

Second example.

pwg reference <ls>MBH. 18, 157.</ls> under headword agADa (अगाध).

The index shows 4 437 XVIII 138 166 29, indicating that verse 157 of 18th parvan occurs on page 437 of volume 4.
And indeed we do find अगाध where expected:

ok to proceed

We are ok to go in preparing the pdf pages and appropriate hosting page for the links.
From 28747 matches in 28741 lines for "<ls>MBH. [0-9]+, [0-9]+.</ls>" in buffer: pwg.txt,
all of these should resolve into links when the necessary infrastructure is developed.

[gasyoun]

all of these should resolve into links when the necessary infrastructure is developed

Hurray! Hope Harivamsa follows and Ramayana is waiting in the line. That's amazing to see it come alive.

[funderburkjim]

@Andhrabharati
You mention some mistakes in the mbh_calc scans (Reference)

• In the vol.1 scan, pp.616-7 are repeated.
  • I cannot confirm this. Would you review?
• In vol.3, p.727 is badly cut at the left side.
  • confirmed. Do you have a replacement from some other source?

The other two noted mistakes to be discussed in separate comments below.

[Andhrabharati]

* In vol.3, p.727 is badly cut at the left side.
  
  * confirmed.  Do you have a replacement from some other source?

you may take this-
https://archive.org/details/in.ernet.dli.2015.326584/page/n743/mode/1up

And you would get the missing pp. 498-9 as well from this archive book.

[Andhrabharati]

Hurray! Hope Harivamsa follows and Ramayana is waiting in the line. That's amazing to see it come alive.

@gasyoun

Now that a format is (kind of) known, you may put your 'new' Russian 'team' on Harivamsa and Ramayana indexing.
[Let others also learn doing small works like these; of course, I had already done the Harivamsa after finishing the MBh indexing last month itself.]

[Andhrabharati]

From 28747 matches in 28741 lines for "<ls>MBH. [0-9]+, [0-9]+.</ls>" in buffer: pwg.txt,

I have 1659 occurrences of <ls>MBH\.</ls> <[0-9]+,[0-9]+> in the VN pages and 28825 in the Main pages, in my 'worked' file.

And there are ~7800 series of MBH. x1,y1. x2,y2. occurrences in the main text and ~100 in the VN pages that need to be separately linked.

[funderburkjim]

Vana Parva verses on p. 517

The above reference notes: In vol.1, p.517 v.3100 onward is numbered as v.4000 onward..
The relevant portion of the index:

1	515	III	3030	3059	30
1	516	III	3060	3087	28
1	517	III	3088	3114	27   <<<<  3088 - 3099, 4000 - 4014
1	518	III	4015	4042	28

consistency with PWG

<s>upAkftya samAhftya</s> <ls>MBH. 3, 3098.</ls> under kar
And consistent with page 517 3088 - 3099 (see previous image)

<s>mOrvIkftakiRO (BujO)</s> <ls>MBH. 3, 4008.</ls> under kiRa.
And consistent with page 517, verses 4000 - 4014.

These examples suggest that the verse numbering in PWG (Vana Parva) is consistent with that in this
mbh_calc_1.pdf. But see next for likely exception.

[funderburkjim]

Three variances to above

3 matches for "<ls>MBH. [3], 3[1-9][0-9]\{2\}.</ls>" in buffer: pwg.txt
These verses (in 3rd parvan) would be in the range 3100-3999. By the above, there should be none.
So these are a puzzle. One turns out to be a typo. The other two are unresolved.

Under ftu, p. 1-1053, 11 lines from bottom
<ls>MBH. 3, 3402.</ls> {#f\tuH svABAvikaH strIRAM rAtrayaH zoqaSa#}

Under garhaRIya, p. 2-0708, 8 lines from bottom
<ls>MBH. 3, 3888.</ls> {#garhaRIyAnyaTA Bavet#}

Under x, p. 6-0401
<ls>MBH. 3, 3995.</ls>
This is a typo: should be 9995.

[Andhrabharati]

Just like to tell that the excel file has one additional column with some details, as Remarks.

This column is not copied in the txt file.

[And I seem to have done few 'unwanted' corrections in the 'To' verse numbers (trying to 'correct' them properly); all such can be reverted to the 'book numbers' looking at the Remarks column, or might even be completely ignored (as they can be derived from the next page starting verse number).]

[Andhrabharati]

These verses (in 3rd parvan) would be in the range 3100-3999. By the above, there should be none.
So these are a puzzle. One turns out to be a typo. The other two are unresolved.

Under ftu, p. 1-1053, 11 lines from bottom <ls>MBH. 3, 3402.</ls> {#f\tuH svABAvikaH strIRAM rAtrayaH zoqaSa#}

This is a printo for 1,3402. And the quoted text does not refer to MBH, but to M. 3,46.

Under garhaRIya, p. 2-0708, 8 lines from bottom <ls>MBH. 3, 3888.</ls> {#garhaRIyAnyaTA Bavet#}

This is a printo for 5,3888.

[gasyoun]

printo

print error?

I had already done the Harivamsa after finishing the MBh indexing last month itself.

So Ramayana left?

[Andhrabharati]

Just did a quick search, to see if there are any more in the range MBH. 3,3100-3999; and found that <L>44247 (pAtAla) has two such.

<ls>MBH.</ls> <3,3547>. fgg. <3552>

[Andhrabharati]

And here are the pages, where some 'jumpimg' of numbering is seen and marked in my excel file-

1.517	III,3088-4014	27	31xx > 40xx : 900
1.520	III,4073-5002	30	41xx > 50xx : 900
1.524	III,5092-6021	30	51xx > 60xx : 900
1.527	III,6082-7011	30	61xx > 70xx : 900
1.530	III,7071-8099	29	71xx > 80xx : 900
1.561	III,8870-9999	30	89xx > 99xx : 1000
1.569	III,10086-11114	29	101xx > 110xx : 900
1.572	III,11071-10294	24	111xx > 102xx : -900
2.36	IV,958-986	29	94x > 95x :10
2.84	IV,2323-2360	28	233x > 234x : 10
2.124	V,990-1019	30	995 > 990 : -5
2.149	V,1578-1598	21	1579 >1578 : -1
2.307	V,6083-7012	30	61xx > 70xx : 900
2.480	VI,4278-4312	30	4286 > 4290 : 5
2.657	VII,3493-3527	30	3505 > 3510 :10
3.34	VIII,927-956	30	928 > 928 : -1
3.250	IX,2015-2049	29	2035 > 2040 : 5
3.374	ΧΙΙ,225-253	29	221 > 225 : 4

The data is slightly reformatted, for quick comprehension.

[Andhrabharati]

printo

print error?

I remember seeing this 'printo' used somewhere long back.

As it rhymed with the popular 'typo' (typographical error in typed documents), I've been using this 'printo' as a short form for 'print error' (for errors in printed books) since then.

[funderburkjim]

Regarding the other 'jumps', I also made a list of 'verse gaps', based on your pdf remarks:
I haven't tried to identify pwg instances of the other verse gaps except that at vol 1, p. 517

gaps in verses. From MBh.Calc.ed.index.Vols.1-4.xlsx

1	517	III	3088	3114	27
<<<<  3088 - 3099, 4000 - 4014  verse gap: 3100 - 3999
1	518	III	4015	4042	28

1	520	III	4073	4102	30
<<<< 4073 - 4099, 5000 - 5002   verse gap: 4100 - 4999
1	521	III	5003	5032	30

1	524	III	5092	5121	30
<<<< 5092 - 5099, 6000 - 6021   verse gap: 5100 - 5999
1	525	III	6022	6051	30

1	527	III	6082	6111	30
<<<< 6082 - 6099, 7000 - 7011   verse gap: 6100 - 6999
1	528	III	7012	7041	30

1	530	III	7071	7099	29
<<<< 7071 - 7099, ---           verse gap: 7100 - 7999
1	531	III	8000	8029	30

1	561	III	8870	8899	30
<<<< 8870 - 8899                verse gap: 8900 - 9899
1	562	III	9900	9927	28

2	35	IV	919	947	29
<<<< 919 - 947,                 verse gap: 948 - 957
2	36	IV	958	986	29

2	83	IV	2293	2322	30
Correction: 2322 -> 2321
2	84	IV	2323	2360	28
Correction: 2323 -> 2322
233x > 234x : 10
Verse gap: 2330 - 2339

2	148	V	1556	1578	23
2	149	V	1578	1598	21
1579 >1578 : -1
Page 148 ends with 1575 and 3 more verses, thus 1575-1578,
Page 149 begins with 2 verses then 1580, thus 1578-1580.
So there is confusion over verse 1578.

2	307	V	6083	7012	30
61xx > 70xx : 900
<<<< 6083 - 6099, 7000 - 7012   verse gap: 6100 - 6999


2	480	VI	4278	4312	30
4286 > 4290 : 5
<<<< 4278 - 4285, 4290 - 4312.  verse gap: 4286 - 4289

2	657	VII	3493	3527	30
(+ 3493 30) = 3523
3505 > 3510 :10
<<<< 3493 - 3504, 3510 - 3527  verse gap: 3505 - 3509

3	250	IX	2015	2049	29
2035 > 2040 : 5
<<<< 2015 - 2034, 2040 - 2049.   verse gap 2035-2039

[funderburkjim]

Page changes

Added 3.498, 3.499 from archive.org version.
Replaced 3.727 from archive.org version.

Also corrected a file-name error in pdf page renaming.

mv pdfpages/mbhcalc_3.502.pdf pdfpages/mbhcalc_3.500.pdf
 . . .
mv pdfpages/mbhcalc_3.861.pdf pdfpages/mbhcalc_3.859.pdf

[funderburkjim]

Links now active for PWG

Example: https://sanskrit-lexicon.uni-koeln.de/simple/pwg/guru

And clicking on the link:

[Andhrabharati]

glad to see this.

you may link the pwk citations also, alongwith mw.

[funderburkjim]

implementation

The url form of the display is https://sanskrit-lexicon-scans.github.io/mbhcalc/?p.v where

• p is the parvan number (1 to 18) and
• v is the verse number.

This displays the pdf of the page containing the given verse for the given parvan.

All the 3000+ page pdfs are in the repository https://github.com/sanskrit-lexicon-scans/mbhcalc.
The display is generated by the web page index.html.

[funderburkjim]

MW citations

There are at least some MW citations of the form <ls>MBh. q,v</ls>. (One such is under 'guru')
Where 'q' is a roman numeral for the parvan.
Such MW citations also have active links in the displays.

[funderburkjim]

PW(K) citations

AFAIK, the PWK MBH citations are to the BOMBAY edition of Mahabharata. They have the form 'MBH. x,y,z' (Three numbers).
These are NOT resolved as links to Calcutta edition.

Not sure if there other PWK citations to Calcutta edition of MBH.

[funderburkjim]

known work to do

• Info -- There is an 'info' link in the MBH verse display. It needs to be written. Maybe someone else could do this?
  • It should provide information on the source of the pdf images
  • And other information that a user might find of interest.
• The 19th parvan. As mentioned in a previous comment, this is stated as HarivaMSa. Probably marked as <ls>Hariv. v</ls>.
• Cleanup of PWG ls references to MBH that are currently garbled or otherwise inaccessible as links.

I'll open other issues for the 2nd and 3rd items. Hope @gasyoun crew will be able to help with the verse-page index for Harivamsa. Probably best for me to do the 3rd item.

[KateRusse]

reality check.

The first MBH reference in pwg is <ls>MBH. 1, 2523.</ls> under headword aMSa. The relevant line of index file is 1 92 I 2513 2541 29 which says in volume 1, page 92 contains parvan 'I' (= 1) verses from 2513 to 2541 (29 verses).

Hello! Thank you, but could you please explain how the number of the parvan is established?

[funderburkjim]

parvan-number correspondence

1 Ādiparva
2 Sabhāparva
3 Vanaparva
4 Virāṭparva
5 Udyogaparva
6 Bhīṣmaparva
7 Droṇaparva
8 Karṇaparva
9 Śalyaparva
10 Sauptikaparva
11 Strīparva
12 Śāntiparva
13 Anuśāsanaparva
14 Āśvamedhikaparva
15 Āśramavāsikamparva
16 Mausalaparva
17 Mahāprasthānikaparva
18 Svargārohanikaparva

We are using a 4-volume pdf for this Calcutta edition. The parvan names and numbers can be seen on title page.

• volume 1 (alternate link to volume 1 titlepage)
• volume 2
• volume 3
• volume 4

Note that in the 4th volume, there is a 19th parvan for Harivansha.

Note that this numbering is consistent with https://en.wikipedia.org/wiki/Mahabharata.

@KateRusse Do these comments answer your question?

[KateRusse]

parvan-number correspondence

1 Ādiparva
2 Sabhāparva
3 Vanaparva
4 Virāṭparva
5 Udyogaparva
6 Bhīṣmaparva
7 Droṇaparva
8 Karṇaparva
9 Śalyaparva
10 Sauptikaparva
11 Strīparva
12 Śāntiparva
13 Anuśāsanaparva
14 Āśvamedhikaparva
15 Āśramavāsikamparva
16 Mausalaparva
17 Mahāprasthānikaparva
18 Svargārohanikaparva

We are using a 4-volume pdf for this Calcutta edition. The parvan names and numbers can be seen on title page.

* [volume 1](https://github.com/sanskrit-lexicon-scans/mbhcalc/blob/main/pdfpages/mbhcalc_1.000a.pdf)  ([alternate link to volume 1 titlepage](https://sanskrit-lexicon-scans.github.io/mbhcalc/pdfpages/mbhcalc_1.000a.pdf))

* [volume 2](https://github.com/sanskrit-lexicon-scans/mbhcalc/blob/main/pdfpages/mbhcalc_2.000a.pdf)

* [volume 3](https://github.com/sanskrit-lexicon-scans/mbhcalc/blob/main/pdfpages/mbhcalc_3.000a.pdf)

* [volume 4](https://github.com/sanskrit-lexicon-scans/mbhcalc/blob/main/pdfpages/mbhcalc_4.000a.pdf)

Note that in the 4th volume, there is a 19th parvan for Harivansha.

Note that this numbering is consistent with https://en.wikipedia.org/wiki/Mahabharata.

@KateRusse Do these comments answer your question?

Yes, thank you very much!

[gasyoun]

And here are the pages, where some 'jumpimg' of numbering is seen and marked in my excel file-

In the original printed book?

I remember seeing this 'printo' used somewhere long back.

Must be Indian English ))

Example: https://sanskrit-lexicon.uni-koeln.de/simple/pwg/guru

Hurray, the Mahabharata links are working! @drdhaval2785 what about the highlighter script that show us the line we actually search for?

may link the pwk citations also, alongwith mw.

Yes, hope we are heading there.

AFAIK, the PWK MBH citations are to the BOMBAY edition of Mahabharata.

@Andhrabharati ?

https://sanskrit-lexicon-scans.github.io/mbhcalc/?p.v

I would propose we get rid of the ? with an Apache rewrite?

crew will be able to help with the verse-page index for Harivamsa

Sure, let @KateRusse take an eye on it. Do you understand what is to be done?

[funderburkjim]

get rid of the ? with an Apache rewrite?

This could be done on the Cologne server, but the current hosting on Github does not support Apache rewrites.

[funderburkjim]

let @KateRusse take an eye on it.

Great! I'll explain what to do in next issue (#49).

[Andhrabharati]

And here are the pages, where some 'jumpimg' of numbering is seen and marked in my excel file-

In the original printed book?

Yes, and it was Hermann Jacobi who has first commented on the point way back in 1903; explaining the total verse number difference in the 3rd parva wrt other editions, counting to over 5000.

The same (count difference) has also been mentioned by the BORI editor in the critical edition.

I remember seeing this 'printo' used somewhere long back.

Must be Indian English ))

could be, as I started looking at "foreign" books only recently.

And this is not listed anywhere 'publicly'.

Example: https://sanskrit-lexicon.uni-koeln.de/simple/pwg/guru

Hurray, the Mahabharata links are working! @drdhaval2785 what about the highlighter script that show us the line we actually search for?

That script works only on the digital texts, not on images.

may link the pwk citations also, alongwith mw.

Yes, hope we are heading there.

There are only a few in pwk, as @funderburkjim correctly noticed. My point was to link them in the same spree.

AFAIK, the PWK MBH citations are to the BOMBAY edition of Mahabharata.

@Andhrabharati ?

Nothing to differ; the pwk biblio entry (vol. 1) itself clearly mentions it.
[¤MBh. = ¤Mahābhārata, citirt nach Parvan, Adhyāya und Śloka der Bomb. Ausg. Die ältere Calcuttaer Ausg. mit zwei Zahlen wird nur dann angeführt, wenn sie eine abweichende Lesart bietet.]

And once the Bomb. ed. got published, attention was shifted to it, as it contained a short (?) commentary as well by Nilakantha.

It won't be out of context to note that MW has heavily picked up all his citations from pwg and hence it has majority of Calc. ed. MBh.

[funderburkjim]

There are only a few in pwk,

@Andhrabharati how do we recognize these? Would you give an example where PWK references the Calcutta edition?

[Andhrabharati]

PW(K) citations

AFAIK, the PWK MBH citations are to the BOMBAY edition of Mahabharata. They have the form 'MBH. x,y,z' (Three numbers). These are NOT resolved as links to Calcutta edition.

Not sure if there other PWK citations to Calcutta edition of MBH.

Even you yourself had the clue, @funderburkjim !

There are 346 places without a dot ending x,y and 137 places with a dot ending x,y., as against 198 of x,y,z and 2327 of x,y,z.

[pl. look at the pw_AB_08.txt file made by you from my file, in November.]

[funderburkjim]

cleanup of pwg MBH instances

The work mentioned in a previous comment .

[funderburkjim]

The cleanup work done in mwg_ls2/mbh directory.

Before the work, there were found 29000+ MBH verses with active links and 8700+ malformed MBH references. After the changes, there are now found 55000+ verses with active links, and 27 known malformed links (these are listed in the readme). About 9000 lines of pwg.txt were changed (less than 1% of the current lines).

The table of 'NORMAL' link types in the readme file shows what are considered as 'normal' ls references to MBH.

Next related improvements to do:

• the links in PWK to MBH calcutta, as identified in comment above
• Similar improvement in PWG markup and display code for HARIV verse links.

[Andhrabharati]

great; so this largest cited MBh. in PWG can now be taken as fully linked.

and let me see what those 27 malformed ones are, mentioned to be in the readme.

[Andhrabharati]

Here is the file to check and act on the 27 cases-
PWG MBh. abnormal cases.txt

And probably some of these could be linked to the resp. pages of the particular volumes, even if they are not pointing to any text verses. If decided to do so and want my help, they (page numbers) could be provided in no time.

[gasyoun]

not pointing to any text verses

Yes, they should be linked as well.

they (page numbers) could be provided in no time.

I ask, not sure if that's enough.

[Andhrabharati]

Need Jim's option, not yours @gasyoun !

[funderburkjim]

PWK Calc. ed. links resolved

As noted above, the links to Calcutta edition of Mahabharata in PWK are identify as, for example
<ls>MBH. 6,352.</ls>, and are already handled properly:
cf. https://sanskrit-lexicon.uni-koeln.de/simple/pw/aDivAjyakulAdya

There are roughly 500 such references. By contrast there are 2500+ references to Bombay edition
e.g. <ls>MBH. 14,38,2.</ls>> https://sanskrit-lexicon.uni-koeln.de/simple/pw/akArpaRya/

Thus nothing more to do in that regard.

[funderburkjim]

PWG.MBh.abnormal.cases.txt

Only 1 of 27 is identified as 'linkable'.
<ls>MBH. Bd. III, S. 818, Z. 5. u. 4, v. u.</ls> under headword aTarvaSiras is identfied as '3, 12864'

An attempted 'obvious' change (e.g. <ls>MBH.</ls> <ls n="MBH. 3, 12864">Bd. III, S. 818, Z. 5. u. 4, v. u.</ls>
was not properly converted to a link in the display program.

Current opinion: Not worth the trouble to modify the display program to handle this isolated case.

[funderburkjim]

Remains to deal with Calcutta edition Hariv. links in PWG, discussed in #49

[Andhrabharati]

PWG.MBh.abnormal.cases.txt

Only 1 of 27 is identified as 'linkable'. <ls>MBH. Bd. III, S. 818, Z. 5. u. 4, v. u.</ls> under headword aTarvaSiras is identfied as '3, 12864'

An attempted 'obvious' change (e.g. <ls>MBH.</ls> <ls n="MBH. 3, 12864">Bd. III, S. 818, Z. 5. u. 4, v. u.</ls> was not properly converted to a link in the display program.

Current opinion: Not worth the trouble to modify the display program to handle this isolated case.

@funderburkjim

Sorry for coming back to a closed issue and posting.

Pl. see my observation as posted at #49 (comment)

Just spent few minutes and found ~1500 places where the [a-z].</ls> is at the end, the MBh. link is NOT active.

So I would venture saying that you need to look at these and do some jugglery, instead of leaving as isolated case(s).
[The above abnormal case is part of this lot]

And here is the list I got, with a quick regex search-
PWG missing MBh links.txt

[funderburkjim]

@Andhrabharati Yes, I've noticed similar MBH misses while working with HARIV. Will focus on missing MBH in a separate issue.