Upgrade StorageHow to convert from Biblioscape (with notes)
In the hopes of helping others who might be trying to move large databases (with note cards) from Biblioscape to Zotero, I thought I'd post the steps that I took.
Introductory comments:
1) The process of exporting my notes and getting them attached to the right references was easier for me than it might be for others, because I had included a numerical tag in each note that designated the reference to which it was attached. (This was a remnant of using the wonderful but now obsolete Papyrus, which gave each reference a unique reference number that was also included in each attached note card.) My notes were also organized mostly by author/citation. I never used Biblioscape's linking function, so I didn't have to worry about loosing this data. If this is the way that you have connected your notes to references, then you will have to do an export that includes the reference in the notecard, or develop some other strategy for linking note cards to their originating reference. You are on your own for figuring this one out!
2) On a similar note, I hadn't yet done much with Biblioscape's categories. I wasn't particularly concerned about loosing what I had. If you are, you might need to come up with some extra tricks.
3) I also did not have many keywords attached to notes. The notes that I had imported from Papyrus (via Scholar's Aid, actually -- I've been at this for awhile!) simply had keywords added to the bottom of each note's text.) So I didn't have the problem of trying to convert Biblioscape note-card keywords into tags. Once again, I can't help you here.
4) Exporting as RTF was the only option for me, because my database had become filled with a variety of fonts and has a lot of foreign language characters in it (German and French). Any other kind of export was filled with HTML font tags. This is one of the many reasons that I wanted to leave Biblioscape, actually. The hard formatting of fonts (the inability to turn everything into easily exportable text) drove me crazy.
5) I had to import into an older version of Zotero (1.0.10 on Firefox 3.0.18) to preserve the line breaks in my notecards. I then upgraded to 2.0 and the line breaks were preserved. (See http://forums.zotero.org/discussion/11487/linebreak-problem-importing-ris-notes/) The developers had, at one point, tweaked the RIS translator to make sure that line breaks were preserved, but this got lost in the upgrade process. It may come back one day...
OK. Here we go.
Steps within Biblioscape:
1) Export your references in RIS format. (Click on your main References folder, Choose File, Export. Then pick "RIS -- Reference Manager" in the "Select the type of tagged ASCII file" menu. Give your file a name in the "Save file as" box, and click Start.)
2) In the Notes module, remove any sub-folders. If you don't do this, when you export the notes, these sub-folders will all be grouped together at the top of the file instead of in front of their child notes. The child notes do get grouped together as you had them organized into folders (a huge relief for me). But as far as I could figure out, the only folder structure that you can actually preserve is the list of main folders in the far left pane (the "Projects" pane). To do this you need to export each one individually, saving them as separate RTF files. (See below.)
3) Right click on the folder of Notes that you want to export in the Project column. Chose File, Export
4) Pick "Delimited / Spreadsheet / DB Table, and click Start
5) Pick .rtf
6) Pick ASCII (MS_DOS). (Note: this is what worked for me. Results may vary.)
7) Tick "Include column titles"
8) Choose Next through the "custom data formats" screen
9) Right click in the "Field Name" box and unselect all
10) Pick only the fields that you want to export and hit Next. (For reasons mentioned above, I had no keywords attached to notes, so I only exported "Title" and "Note.")
11) Next through custom header screen
12) Pick "export data only" and "tabular form." Click Next.
13) Give a title and location for your file and hit Execute.
From RTF to RIS:
1) Open the RTF in Word (for whatever reason I found Word for Windows much more stable than Word for Mac, which kept crashing with the large RTFs) and clean up any problems you see (extra spaces etc.)
2) Select the far left column. Search " Title" [one space before the N] and replace with "N1 " [2 spaces after the 1]
3) In the same column, select " Note" [one space before N] and replace with nothing
4) Partial instructions: If you've exported keywords, figure out this step. I didn't have to do it. But I will tell you that the RIS tag for a Zotero tag would be "KW - " [two spaces before and one space after the hyphen] and each one would have to be on its own line.
5) Select one row of the table first, then choose Table, Select Table. Convert the table to text using a hyphen as the separator.
6) Save the file as text. (I used Unicode UTF-16, because this worked best to preserve the German and French characters in my database. UTF-8 did the trick for Zotero 2.0 on a Mac, but I had to use UTF-16 for the older version on Word for Windows in XP. A mystery to me, but that's what worked.)
7) Open the text file and remove any strange characters at the beginning of the file (like "") and footers (like "SMExport 4.77") at the bottom.
Combing References and Notes to make your import file:
1) Open up the RIS file of your exported references. (Remove strange characters as mentioned above.)
2) Matching up the correct notes with the correct references, insert all of the notes for each reference before the closing "ER -" tag. Each "N1" note card needs to start on a new line.
Voila. Your file should now be ready to import into Zotero.
Some glitches that might arise:
1) I had to do some searching and replacing of foreign characters, because a few different ways of coding them had crept into my database. (A consequence of Biblioscape preserving every font/language coding known to man?)
2) Books that did not have full publication data (e.g. where the city was missing) exported as Document. I figured this out late and had to make the changes in Zotero after exporting. Looking first and doing some searching and replacing might be quicker.
3) You will have the usual problems with RIS imports as described in other postings -- such as the editor of edited collections being designated as "Contributor." Barring new solutions to these problems, some manual clean up seems unavoidable. Such are the costs of database conversion.
Needless to say, these instructions are what worked for me. I am just an historian (not a programmer) and I can't explain any of these steps in any more detail than I have provided here. But I hope that this might save someone out there some time and experimentation. Happy converting!
A.
Introductory comments:
1) The process of exporting my notes and getting them attached to the right references was easier for me than it might be for others, because I had included a numerical tag in each note that designated the reference to which it was attached. (This was a remnant of using the wonderful but now obsolete Papyrus, which gave each reference a unique reference number that was also included in each attached note card.) My notes were also organized mostly by author/citation. I never used Biblioscape's linking function, so I didn't have to worry about loosing this data. If this is the way that you have connected your notes to references, then you will have to do an export that includes the reference in the notecard, or develop some other strategy for linking note cards to their originating reference. You are on your own for figuring this one out!
2) On a similar note, I hadn't yet done much with Biblioscape's categories. I wasn't particularly concerned about loosing what I had. If you are, you might need to come up with some extra tricks.
3) I also did not have many keywords attached to notes. The notes that I had imported from Papyrus (via Scholar's Aid, actually -- I've been at this for awhile!) simply had keywords added to the bottom of each note's text.) So I didn't have the problem of trying to convert Biblioscape note-card keywords into tags. Once again, I can't help you here.
4) Exporting as RTF was the only option for me, because my database had become filled with a variety of fonts and has a lot of foreign language characters in it (German and French). Any other kind of export was filled with HTML font tags. This is one of the many reasons that I wanted to leave Biblioscape, actually. The hard formatting of fonts (the inability to turn everything into easily exportable text) drove me crazy.
5) I had to import into an older version of Zotero (1.0.10 on Firefox 3.0.18) to preserve the line breaks in my notecards. I then upgraded to 2.0 and the line breaks were preserved. (See http://forums.zotero.org/discussion/11487/linebreak-problem-importing-ris-notes/) The developers had, at one point, tweaked the RIS translator to make sure that line breaks were preserved, but this got lost in the upgrade process. It may come back one day...
OK. Here we go.
Steps within Biblioscape:
1) Export your references in RIS format. (Click on your main References folder, Choose File, Export. Then pick "RIS -- Reference Manager" in the "Select the type of tagged ASCII file" menu. Give your file a name in the "Save file as" box, and click Start.)
2) In the Notes module, remove any sub-folders. If you don't do this, when you export the notes, these sub-folders will all be grouped together at the top of the file instead of in front of their child notes. The child notes do get grouped together as you had them organized into folders (a huge relief for me). But as far as I could figure out, the only folder structure that you can actually preserve is the list of main folders in the far left pane (the "Projects" pane). To do this you need to export each one individually, saving them as separate RTF files. (See below.)
3) Right click on the folder of Notes that you want to export in the Project column. Chose File, Export
4) Pick "Delimited / Spreadsheet / DB Table, and click Start
5) Pick .rtf
6) Pick ASCII (MS_DOS). (Note: this is what worked for me. Results may vary.)
7) Tick "Include column titles"
8) Choose Next through the "custom data formats" screen
9) Right click in the "Field Name" box and unselect all
10) Pick only the fields that you want to export and hit Next. (For reasons mentioned above, I had no keywords attached to notes, so I only exported "Title" and "Note.")
11) Next through custom header screen
12) Pick "export data only" and "tabular form." Click Next.
13) Give a title and location for your file and hit Execute.
From RTF to RIS:
1) Open the RTF in Word (for whatever reason I found Word for Windows much more stable than Word for Mac, which kept crashing with the large RTFs) and clean up any problems you see (extra spaces etc.)
2) Select the far left column. Search " Title" [one space before the N] and replace with "N1 " [2 spaces after the 1]
3) In the same column, select " Note" [one space before N] and replace with nothing
4) Partial instructions: If you've exported keywords, figure out this step. I didn't have to do it. But I will tell you that the RIS tag for a Zotero tag would be "KW - " [two spaces before and one space after the hyphen] and each one would have to be on its own line.
5) Select one row of the table first, then choose Table, Select Table. Convert the table to text using a hyphen as the separator.
6) Save the file as text. (I used Unicode UTF-16, because this worked best to preserve the German and French characters in my database. UTF-8 did the trick for Zotero 2.0 on a Mac, but I had to use UTF-16 for the older version on Word for Windows in XP. A mystery to me, but that's what worked.)
7) Open the text file and remove any strange characters at the beginning of the file (like "") and footers (like "SMExport 4.77") at the bottom.
Combing References and Notes to make your import file:
1) Open up the RIS file of your exported references. (Remove strange characters as mentioned above.)
2) Matching up the correct notes with the correct references, insert all of the notes for each reference before the closing "ER -" tag. Each "N1" note card needs to start on a new line.
Voila. Your file should now be ready to import into Zotero.
Some glitches that might arise:
1) I had to do some searching and replacing of foreign characters, because a few different ways of coding them had crept into my database. (A consequence of Biblioscape preserving every font/language coding known to man?)
2) Books that did not have full publication data (e.g. where the city was missing) exported as Document. I figured this out late and had to make the changes in Zotero after exporting. Looking first and doing some searching and replacing might be quicker.
3) You will have the usual problems with RIS imports as described in other postings -- such as the editor of edited collections being designated as "Contributor." Barring new solutions to these problems, some manual clean up seems unavoidable. Such are the costs of database conversion.
Needless to say, these instructions are what worked for me. I am just an historian (not a programmer) and I can't explain any of these steps in any more detail than I have provided here. But I hope that this might save someone out there some time and experimentation. Happy converting!
A.