- News & Updates
- Using Digital Humanities to Develop Citizen Scholars
- Digital Humanities in the Classroom
- Treebanking Pilots
- Uploading Google Spreadsheet Data As Annotations
- Telécharger dans Perseids des annotations avec des Feuilles de calcul Google
- Creating a New Treebank Annotation
- Instructions for Using Templates for Treebanking with Students
- Adding Standard References to Treebank Files
- Instructions for Annotating Quotation Sources in Athenaeus
- Leipzig Summer School Instructions
- Bodin Spring 2014 Perseids Instructions
- Iliad Translation Alignment Instructions
- Setup Greek Keyboard
- Windows Instructions at http://www.jcu.edu/language/llc/keyboard-setup-greek.htm
- Review CTS & CITE Syntax
- See http://shot.holycross.edu/~nsmith/code/cite-doc/docs/index.html and/or the cite-ov.pdf in the epifacs Google Drive folder.
- Download and install oXygen or other XML editor (Perseus has student licenses for oXygen — see Lisa Cerrato for details)
- Create Google email address (if you don’t have one already) and send it to Bridget Almas
Identifiers and Namespaces
We are using CTS URNs as identifiers for the inscriptions, their transcriptions and translations. The syntax is as follows:
We are going to try to use the common abbreviation for the collection in which the inscription can traditionally be found as the namespace. This might not always work, but we will adjust as needed as we go. So for examples, inscriptions found in Inscriptiones Graecae will use the namespace ‘ig’; those from CIL will use ‘cil’; and so forth.
The inscription and group ids are restricted to characters in the ASCII character set, excluding . / , :
For Inscription and group ids we will try to use the group and inscription identifiers from the source collection. If the group identifier normally includes any of the restricted characters (./,:) , they will be replaced with underscores (_). Parentheses () will also be replaced with a single underscore for the opening parentheses, and only one at the end if it is followed by additional characters. Spaces will be eliminated.
Version ids apply to each edition and translation and will be assigned by the Sosol tool.
Sample CTS work urns:
For IGVII 2543-2545 : urn:cts:ig:vii.2543-2545
For CIL 06, 05326 (1) : urn:cts:cil:06.05326_1
We are using CITE Collection URNs as identifiers for the images. The syntax is:
The collection namespace is ‘epifacs’.
The group is ‘epifacsimg’
The ID and VERSION components will be sequentially increasing numbers assigned automatically by the collection editor tool.
Example CITE urns for images:
- Check the TextInventory in the epifacs.xml file in the epifacs Google Drive folder to make sure the inscription is not already listed (if it is, skip the remaining steps)
- Identify the collection to which the inscription belongs
- If there is not a textgroup entry in the epifacs.xml file for that collection, add a new one. Contact Bridget Almas if you are unsure of what the collection projid should be.
- Add a work entry for the inscription under the appropriate textgroup. Contact Bridget Almas if you are unsure of what the work projid should be.
Images to be inventoried are in the shared netapp drive in the epifacs subdirectory. (See instructions on mapping the netapp drive). Or you can access them in the Image Service directly by using the links provided at http://services.perseus.tufts.edu/epifacs/mcbimages.html . (Once they’ve been inventoried you can will also be able to browse them via the “Available Collections” link at http://sosol.perseus.tufts.edu/collections/ as described further in the section on browsing below).
First, identify the prefix to use with the image name, by the directory in which in can be found. Because the directory names are long, and because many images have duplicate file names, when the images are converted to tifs, a new file name was created by prefixing the original filename with an abbreviation for the directory in which it was founds. The directory name to prefix mappings are below:
Baths of Diocletian (epigraphical museum) : BD
Cave of the Sibyl : CS
Firenze, Archaeological Museum : FAM
Naples, Archaeological Museum : NAM
Naples misc : NAP
Naples, San Gaudioso + catacombs s.Gaudioso : NSG
Nedymos : NED
Ostia Antica : OA
Palazzo Massimo : PM
Pompeii : PEI
Pozzuoli : POZ
Rome, Capitoline Museum : RCM
Rome, churches : RCH
Rome, forum : RF
Rome, San Clemente : RSC
St-Peter’s : SP
Thasos Sarcophagus : TS
Trevi fountain : TF
Vatican Museums : VM
So, for example, the file DSC_0832.JPG from the Baths of Diocletian (epigraphical museum) directory has been converted to TIF format under the name BD.DSC_0832.
Using the correct name for the TIF version of the image is important.
Once you have the name for the TIF version of the image, you can view it in the IIP Image Service at the following URL, replacing IMAGENAME with the correct name.
So, for example, to view BD.DSC_0832, the link is
Next, try to identify the inscription to which the image belongs by its URN (See above under “Identifiers and Namespaces” for instructions on determining the CTS urn for the inscription.
Adding a new Image to the Inventory
To add a new image to the collection inventory, go the following url:
And click on the “Add to Collection Inventories” Link:
If you have not yet logged in to the collection editor, you will be prompted to do so:
Click the “Click Here to Authorize” link, and login with your google account. You’ll be returned to the page and it will be ready for you to add a new inventory entry. The CITE URN for the new image will already be populated for you:
Add as much of the requested information as you are able. It’s okay if you can’t complete everything initially.
- Image Name: this is the name of the image, created as described above by prefixing the original file name with the abbreviation for the directory followed by a . and dropping the extension. (E.g. BD.DSC_0832.)
- Rights: just accept the default of the Creative Commons Attribution license
- Caption: If you can, supply a caption for the image
- Title: if you can, supply a title for the image
- Description: If you can, supply a description for the image.
- Creator (Image): accept default (for now just Marie-Claire we will add more to the list later if we need to)
- Date (Image): This should be the date the picture was taken, if known
- Edited By: Should be prepopulated with your name as you have indicated it in your Google account profile
- Work URN: This should be the URN for the inscription depicted if you know it. If it’s one already in the text inventory, then use the identifier from there. Otherwise, try to create the work urn following the instructions outlined in “Identifiers and Namespaces above”.
- Work Creator: This should be the original creator of the inscription, if known. Unknown is okay.
- Work Date: This should the original date the inscription itself was made, if known.
Click “Submit” to store your entries in the database. You can also click “Save” which will save your entries temporarily in the browser’s local store but it won’t submit them to the server. This is good to do if you are working with a poor internet connection or otherwise have to leave the page before are done and don’t want to lose your work. But note that you must click “Submit” if you want your work to be preserved in the inventory actually make it to the server and the database.
Browsing and Editing the Image Inventory Entry
You can browse the image inventory and edit existing entries in it by clicking on the “Available Collections” link at http://sosol.perseus.tufts.edu/collections/. This will bring you to the list of Available Collections (currently just the Epifacs Image collection):
Click the collection urn link, and you will see the list of entries. Wait a minute or two for all of the image names to load next to the URNs:
If the image names don’t load, click the browser refresh button to refresh the page. This list may take a little while to load as the number of images inventoried increases. Eventually paging support will be added.
Click the urn link for the entry you want to view:
You should see the thumbnail for the image, and the data that has been entered in the inventory for that image. The thumbnail is clickable to see a larger view of the image.
If you don’t see a thumbnail, then the problem is likely an incorrectly entered image name.
If you want to make changes or additions to the inventory entry, click the urn link at the top of the page. Note that when you edit an entry, what you are actually doing is making a new version of the item in the collection. The version number is assigned automatically by the collection editor tool. Clicking the link brings you to an entry form for the new version, with the fields all pre-populated from the version you are editing. E.g., if I click the link to edit urn:cts:epifacs:epifacsimg.5.2, the tool will create the next version of this entry, urn:cts:epifacs:epifacsimg.5.3, and so forth.
If you return to the list of images and don’t see the new version you just edited, click Refresh to make sure the page has reloaded.
Working on Texts (Transcriptions, Translations and Commentary)
- Use SoSOL tool at http://sosol.perseus.tufts.edu/sosol to work on transcriptions and translations.
- If the inscription you want to work on is not listed in the SoSOL tool, make sure it is in the epifacs.xml TextInventory (in the epifacs Google Drive folder). If it is, contact Bridget Almas to update the inventory for the Sosol environment. If it is not, follow steps above under Inventorying Inscriptions.
Mapping Images to Text
- Images are currently kept in the images folder of the epifacs Google Drive collection.
- The indexdata spreadsheet in the epifacs Google Drive collection is where we are currently recording information about the images and their mappings to the text (this will be moved to SoSOL or the CITE collection editor interface in the future)
- If the image you want to map is not listed in the indexdata spreadsheet on the img.img and the refidx.idx_entry worksheets and/or isn’t available in the Image Mapping tool at http://services.perseus.tufts.edu/ICT/index.html?w=600&urn=URN (where URN is replaced with the CITE urn of the image), follow the steps above under Inventorying Images.
Using Image Citation Tool to map image coordinates to text coordinates
- Load the image in the web based Image Citaiton Tool at the link indicated on the img.img worksheet of the indexdata spreadsheet
- this should be a link to a url like http://services.perseus.tufts.edu/ICT/index.html?w=600&urn=URN where URN is replaced with the CITE urn of the image
- E.g. for the image with the urn urn:cite:epifacs:epifacsimg.DSC03651 the url is http://services.perseus.tufts.edu/ICT/index.html?w=600&urn=urn:cite:epifacs:epifacsimg.DSC03651
- Highlight the image to be mapped with your mouse, and when you are satisfied with the selection, copy the automatically calculated CITE-URN + ROI from the page to the ‘val’ column of a new row in the refidx.idx_entry tab in the indexdata worksheet.
- Create a CTS urn for the mapped text region in the ref column of the new row in the worksheet.
The image viewer software we’re using (IIP Image) requires that the tiffs be tiled, per the instructions at http://iipimage.sourceforge.net/documentation/images/ .
We can use the ImageMagick tool to identify and convert tools to test and convert the images.
Use identify to check whether a tif is tiled.
The output for an untiled tiff will show information for one layer? only, as in:
DSC03651c.tif TIFF 2592×1944 2592×1944+0+0 8-bit DirectClass 10.86MB 0.000u 0:00.000
Versus output for a tiled tiff which shows information for multiple layers? as in:
DSC03651c.tif TIFF 2592×1944 2592×1944+0+0 8-bit DirectClass 5.9MB 0.000u 0:00.000
DSC03651c.tif TIFF 1296×972 1296×972+0+0 8-bit DirectClass 5.9MB 0.000u 0:00.000
DSC03651c.tif TIFF 648×486 648×486+0+0 8-bit DirectClass 5.9MB 0.000u 0:00.000
DSC03651c.tif TIFF 324×243 324×243+0+0 8-bit DirectClass 5.9MB 0.000u 0:00.000
DSC03651c.tif TIFF 162×121 162×121+0+0 8-bit DirectClass 5.9MB 0.000u 0:00.000
DSC03651c.tif TIFF 81×60 81×60+0+0 8-bit DirectClass 5.9MB 0.000u 0:00.000
I’m not sure how to convert directly from an untiled tiff to a tiled tiff, but I have been able to get the conversion to work by way of jpeg.
Convert to jpeg:
convert DSC03651a.tif -compress jpeg ‘jpeg:DSC03651a.jpeg’
convert DSC03651a.jpeg -define tiff:tile-geometry=256×256 -compress jpeg ‘ptif:DSC03651a_tiled.tif’