Transliteration-related information

[oshfdk]

I'm not sure where and how to report these, I think there is a typo in the transliteration. I've checked what I believe is the latest transliteration file You are not allowed to view links. Register or Login to view.

# Small ring of text around central moon face
<f68r3.22,@Cc>    <!09:30>otcheody.chokchy.okol.cheol.dar.cho.keol.dolaiin.okeol.oly

The first word looks like "otchody" on the scan.

[nablator]

The first word looks like "otchody" on the scan.

And, first line I checked on the same page f68r3, outer ring: sor.cheol.oteo.okeey should be sol.cheor.oteo.choteey as in the reference transliteration (RF).

You are not allowed to view links. Register or Login to view.
You are not allowed to view links. Register or Login to view.

Better use You are not allowed to view links. Register or Login to view., it has less mistakes it seems.

[ReneZ]

Thanks! 
This is the best place for this typo of info.
Version 3b is gradually being updated but not yet published.
Undoubtedly, there are still many uncontested errors in all files.
RF should be closest to the truth.

[ReneZ]

Earlier today I posted a Zip file with short text files, one per page. The file naming convention is one I routinely use, and is based on the page variables available in the IVTFF files.

The variable $Q has a single upper case character identifying the quire, where A=1 and T=20.
The variable $P has a single upper case character identifying the 'page in quire', starting at A and running up to whatever is the limit. For several reasons, some letters may be skipped. However, the $P variable increases as one leafs through the quire.

Page file name EN.txt refers to quire E (5), page-in-quire N, which turns out to be f39v.

This system has several advantages:
- all file names have the same pattern
- you can identify the quire from the file name
- the regexp ??.txt expands to the complete set of pages in the right order

Unfortunately, these names are unfamiliar and not easily mapped to the normal folio names.
Of course, the regexp f*.txt would expand to all pages in the MS using the familiar folio names, but not in the right order.

Perhaps I have a simple script somewhere (227 lines long) that renames all files, but I won't bet on it as I never needed it. It would probably take about 10 minutes of copying, pasting and modifying, which is unfortunately error-prone.

[ReneZ]

Following are three basic shell scripts that allow one to process all pages of a transliteration separately. I have used this for many different purposes, and just creating clean texts per page is one of the simplest applications. The scripts are short enough, I think, to just put them inline here.

The highest level script is called all_sh and this is the command one should type, without any arguments. It also does not need to be changed.

---------------------------------------------------------------------------------------------------------

#/bin/sh
#
# Do something for all pages in the MS
#
many_sh A A B C D E F G H I J K L M N O P
many_sh B A B C D E F    I J K L M N O P
many_sh C A B C D E F G H I J K L M N O P
many_sh D A B C D E F G H I J K L M N O P
many_sh E A B C D E F G H I J K L M N O P
many_sh F A B C D E F G H I J K L M N O P
many_sh G A B C D E F G H I J K L M N O P
many_sh H A B C D E F G H
many_sh I  B C  E F  H I J  L M N
many_sh J A B  D E  G H
many_sh K A B  D E F  H I J
many_sh L A B
many_sh M A B C D E F G H I J K L M N O P Q R S T
many_sh N  B C D                  N O P Q
many_sh O A B C D  F G  I J  L M  O P
many_sh Q A B C D  F G  I J K L
many_sh S A B C D E  H        L M  O P
many_sh T A B C D E F G H I J K L M N O P Q R S T U V W X


---------------------------------------------------------------------------------------------------------

This invokes many_sh which also hardly ever needs to be changed. In the following version it saves time by creating temporary files for one quire.

---------------------------------------------------------------------------------------------------------

#!/bin/sh
#
# Repeat a command for all pages in a quire
# Arg 1 is $Q of the quire
# All further arguments are the $P of the pages in that quire
#
qq=${1}
shift
echo "quire " ${qq}
#
qfile=quire${qq}.txt
ivtt -m1 +Q${qq} RF1a-n.txt >${qfile}
for pp in "$@"
do
  once_sh ${qq} ${pp} 2>/dev/null
done
\rm -f ${qfile}


---------------------------------------------------------------------------------------------------------

Finally, the 'clever stuff' is done in once_sh, which in this case is trivial.

---------------------------------------------------------------------------------------------------------

#!/bin/sh
#
# Run a short script once.
# Arg 1 is the $Q variable for a quire.
# Arg 2 is the $P variable for one page.
# It is assumed that a transliteration file for this one quire was
# already prepared at a higher level.
#
qfile=quire${1}.txt
pfile=${1}${2}.txt
#
ivtt -x8 +P${2} ${qfile} >${pfile}


---------------------------------------------------------------------------------------------------------

[bi3mw]

@Rene:

A note on the files. In "many_sh" the output should not be sent to the null device. No text files are then produced.

Code:

before: ------- for pp in "$@" do   once_sh ${qq} ${pp} 2>/dev/null done after: ------ for pp in "$@" do   once_sh ${qq} ${pp} done

.... and all files must of course be in the path, otherwise there must be "./" before the calls.

[nablator]

Unfortunately, these names are unfamiliar and not easily mapped to the normal folio names.

grep "<! $Q" RF1a-n.txt | sed "s/^<\(.*\)>  *<! $Q=\(.\) $P=\(.\).*$/mv \2\3.txt \1.txt/"

[bi3mw]

Perhaps I have a simple script somewhere (227 lines long) that renames all files, but I won't bet on it as I never needed it. It would probably take about 10 minutes of copying, pasting and modifying, which is unfortunately error-prone.

I have renamed the files with a script. The code must be saved in a file, e.g. "gen_list.sh".

- The file "RF1a-n.txt" must be in the same directory as "gen_list.sh".
- The folder "RF" from You are not allowed to view links. Register or Login to view. must be in the same directory.
- "Your_path" in "gen_list.sh" must be adapted ( reference to directory in which "gen_list.sh" is located )

Code:

#!/bin/sh grep '<! $Q=' RF1a-n.txt >> folio_list.txt awk -F'>' '{print $1}' folio_list.txt >> folio_list2.txt rm ./folio_list.txt cat folio_list2.txt | sed -e 's/^.\{,1\}//' > folio_list3.txt rm ./folio_list2.txt mv folio_list3.txt folio_list.txt cd /Your_path/RF ls  | paste - /Your_path/folio_list.txt | awk 'NF == 2 { print "mv ", $1, $2 } ' | bash rm -f /Your_path/folio_list.txt

  all_pages.zip (Size: 126.24 KB / Downloads: 22)

[ReneZ]

In many_sh only the standard error output is redirected to the null device. The scripts just need to be in the path, and the command needs to be invoked in the directory that has the RF1a-n file. Usually I set up an empty directory where the transliteration file is just a symbolic link.
I used literally these scripts to create the set.

@nablator: brilliant 
I was kind of hoping for a response of this type

[MarcoP]

Hi Rene,
I am playing around with the planet/zodiac matches for f67r2 proposed by oshfdk.
I noticed a typo in ZL3a-n.txt:

<f67r2.53,&Ls> dchdar

should be

<f67r2.53,&Ls> dchodar

IT2a-n.txt has the correct string.