Search for content in message boards

My current method of coping without search/annotation/pattern recognition

Replies: 0

My current method of coping without search/annotation/pattern recognition

Posted: 8 Oct 2012 11:43PM GMT
Classification: Query
There are other ways to do this, and I'm sure there are better ways, but I thought I would describe my current method for being able to organize my matches meaningfully. I only wish I'd done this from the very start. I'm at over 1600 matches now, which of course is wonderful, but even after cutting out the ~400 private trees, it's just a mess to manually organize with any system. And heaven help the person who needs to look for something they weren't checking for when they opened the first thousand matches. (If we're lucky, eventually that's all of us.)

What I'm really hoping is that describing my method here will lead to "wash the car to make it rain" syndrome, and tomorrow Ancestry will debut search/annotation features and something to track recurring surnames not already in your tree. Fingers crossed! :)

This method may sound a bit fiddly, but it only adds about 10-20 seconds to looking over a match, depending on how much you have to type.

Apologies in advance for being Windows-centric in these instructions.


MATERIALS
=========
* Two web browsers (ideally one being Firefox for reasons discussed below)
* Google Docs
* Textpad (Free trials can be downloaded from Textpad.com. However, any robust text editor will do, aka not Notepad. Word processors might have the necessary features; I just don't use them for these kinds of tasks because text editors are lighter/easier.)


PREP
====
Make a form in Google Docs. This form is where you're going to input data from each match. GDocs will then transform it into a spreadsheet. I started out just doing a spreadsheet manually, but scrolling around was getting really unwieldy.

See the attached image for what I have in my form. I have fields for basic information (URL, user name, confidence level, whether there's a Shared Ancestor Hint), a place to note any obvious connections, and places to note surnames in specific geographical areas. (Alas, Ancestry only looks at birth info for locations, but I'll take what I can get.)

You can always update the form with new fields later. If you do, the spreadsheet will show those cells in a different color for matches from before that field was added. That way, it's easy to see who needs to be re-checked for the new data.


PROCESS
=======

1. Open your live GDocs form in one web browser.

I use two browsers because I find it faster to ALT+TAB between windows than to switch to the correct tab when many tabs are open. I also prefer keeping my hands on the keyboard instead of pausing to grab the mouse, so this is what works for me.

2. Open your main list of DNA results in another browser that isn't Google Chrome.

I love Chrome, but it doesn't save web pages as text files. Unfortunately, a week or two ago Firefox stopped working for me when trying to look at match pages. (Every other page on Ancestry and the rest of the Internet is fine. It's a known issue affecting a few people so hopefully it will be sorted soon.) This means I'm stuck with IE. IE isn't quite as good here because you have to remind it every single time to save files as text, but I'm too lazy to download even more browsers, lol.

3. Right-click "Review Match" and open each match in a new tab. (How many tabs to open depends on how many tabs it takes to make your computer grumpy and slow.)

4. Visit each tab and do the following:

a. CTRL+S to save the file as text. Name the file the same as the username.

b. Copy the URL and paste it into the form.

c. Fill in other basic info on form.

d. Examine your match as usual. Switch to the form as needed to fill in any fields particular to your interests.

e. Press submit when done. Click the link to start the form over. On to the next tab and begin again!


SEARCHING
=========

Textpad has great searching features. Press CTRL+F5, point at the directory where you saved all of your match web pages as text files, and type in what you want to look for.

Textpad will then search all of those files for the name or location you entered. It will give you a report of the file name and the line with the info on it. Note the file name (same as the user name), find it in your spreadsheet (that the form automatically created), and follow the link.

I've had a few occasions (one just today) where I noticed that a surname in an area (that I was looking at for other reasons) kept reappearing. This has led to some great discoveries. (It has also caused me to question some previously "verified paper trail" matches as new connections to the same match arose, but I'm not going to get on my soapbox once again about the perils of thinking you've "verified" your DNA match without comparing the actual segments. Ah, restraint! :)


ORGANIZING
==========

Despite my luck earlier today, with so many matches looked at over a period of months, my brain doesn't always notice how often certain surnames outside my tree are recurring. The special fields on my spreadsheet let me check for surname patterns in a few areas. (This is so limited, but it's better than nothing.)


FOR THE DESPERATE AND/OR NERDY
==============================

Another benefit to using Textpad (instead of just searching files in Windows Explorer) is that you can make macros to do little jobs, which brings us to the next bit. This is how I looked at all high confidence matches for recurring surnames.

1. Put copies of all high confidence match text files in a separate directory.

2. Open all of these files (at once) in Textpad and delete the text above and below the list of surnames. (All surnames, not shared surnames.) Search for "Surnames (10 generation pedigree)" to see about where to stop deleting at the beginning, and search for "1st gen2nd" to see where to start chopping at the bottom.

3. Sort each file (F9).

4. The top of the file will now be several blank lines then several lines of individual names. Use SHIFT+DOWN ARROW (or your mouse) to highlight and delete all of that. You are then just left with a list of surnames. (They will have numbers after them, but these can be ignored.)

5. Go to your first file and make the following macro:

a. Put the cursor at the end the first surname.
b. Click the macro recording button (or CTRL+SHIFT+R).
c. Hit TAB.
d. Type in the user name for that file. (It will be visible as the file name.)
e. Move your cursor to the next line.
f. Hit the END button.
g. Click the button to stop recording (or CTRL+SHIFT+R again).

(You don't have to save the macro. It will be there until you record a new one or close Textpad.)

6. Run the macro through the rest of the file by hitting the "play" button (or CTRL+R) over and over. (Or, if you saved the macro, you can tell it to play through the rest of the file.)

7. Redo the macro for each file. (Redo because you have to type a different user name for each file, although you could run the same macro then just press F8 to replace all instances of the old name to the new name.) Run the macro for each file.

Each file now has a list of surnames with the user name after it.

8. Make a new text file.

9. Copy and paste the contents of each file into this file. (CTRL+A, CTRL+C, go to new file, CTRL+V.) I think there's a Unix-like concatenate command for Windows, but this didn't take me very long.

10. Save the new file.

11. Press F9 to sort all of the surnames.

12. Start looking for recurring surnames!

(Update the master file of high match surnames as needed.)

The above sounds like a huge pain (and I know it's a kludgy way to do things - the part of me that used to write perl scripts is weeping in shame), but all in all it probably took me under 20 minutes and I was deeply rewarded. Two surnames beyond my "too common to bother with at this time" threshold came up a few times. I then searched for those names among all my text files and found so many matches to two particular families that I know I must be on to something hot.

I've also used similar methods to look and concatenate for just files with a certain surname to see if named ancestors/locations are recurring.


THE END
=======

Yes, it's absolutely nutso that we don't have even a basic search engine and that some of us are going to such lengths. However, it *has* paid off for me to the point where I wrestle with going back and saving all of the matches from before I started doing this so they will be searched, too. (So far I've only gone back to do the High/Moderate ones. After that my brain shuts down and keeps repeating, "Why am I manually searching in 2012?!")

Anyway, I'm sharing in case it helps any diehards... and also in the hope of rain. :)
Attachments:

Find a board about a specific topic

  • Visit our other sites:

© 1997-2014 Ancestry.com | Corporate Information | Privacy | Terms and Conditions