Search Companion v1.06

Info Pattern Maker

Script editor:

Minimal character number for whole line:

Format options:	Must exist (e) Not output (n) Unique (u) Memo type (m)

Include Title of fields Include URL of source pages

Pattern:
Format:
Comment:

Operation Guide:

Many users wanted to search for info with InfoSeek/FastSeek, but are not able to customize patterns by themselves. Here is an instance to search a familiar website style (paged site) with InfoSeek and Search Companion, and detailed usages and steps for search are provided. For example with the following paged site, the left is the content of the first page, the right is is the content of first message. Now we will search for info of messages in 100 pages.

Page 1:

Message 1:

Page 1

* Message 1......................
* Message 2......................
* Message 3......................
* Message 4......................
* Message 5......................

Page: 1 2 3 4 5 Total: 100

Prev page
ABC Inc.
Address: 1129 Cole Ave, Los Angeles, CA	Zip: 90038	Contact: Mr. William
Tel: (323) 466-9984	Fax: (323) 466-9984
Email: abc@hotmail.com	Site: http://www.abc.com
Overview: XXXXXXXXXX XXXXXXXXXX XXXXXXXXXX Next page

Notice: before the following steps, you must import two jobs: in InfoSeek, select menu: "Edit">>"Job>>"Import", goto "Data" subdirectory in the install directory of InfoSeek, import "page_address.ini" and "page_info.ini" files. After returning to the main interface, two new jobs: "page_address" and "page_info" will be added.

Step 1: Search pages for addresses of messages (with "page_address" job of InfoSeek)

1. Customize "page_address" pattern

With "Address Pattern Maker", copy the source code of the first page to "Page source code" editor. For example:

Page 1
<br><br>
* <a href="http://www.test.com/article.htm?id=1">Message 1......................</a><br>
* <a href="http://www.test.com/article.htm?id=2">Message 2......................</a><br>
* <a href="http://www.test.com/article.htm?id=3">Message 3......................</a><br>
* <a href="http://www.test.com/article.htm?id=4">Message 4......................</a><br>
* <a href="http://www.test.com/article.htm?id=5">Message 5......................</a>

<br><br>
Page:
<a href="http://www.test.com/result.htm?page=1">1</a>
<a href="http://www.test.com/result.htm?page=2">2</a>
<a href="http://www.test.com/result.htm?page=3">3</a>
<a href="http://www.test.com/result.htm?page=4">4</a>
<a href="http://www.test.com/result.htm?page=5">5</a>
Total: 100
<br>

(Page source code)

Select addresses of two different messages, such as: Message 1 and Message 2, copy addresses to "Address sample one" and "Address sample one". For example:

(Address sample one)
(Address sample two)

Click "Address Pattern" button to get the pattern and its format. Open "Options" dialog of InfoSeek (select menu "File">>"Options"), switch to "Pattern" tab, copy data to "page_address" pattern.

Copy page source code: right click on the page and select "View Source" (or select IE menu "View">>"Source"), copy all text in opened Notepad.
Copy address (URL): move mouse over a link, select "Copy Shortcut" from right-click menu.

2. Add pages to search

In "Batch Address Creator", input "The first page address", "The second page address" and "Total page number", click "Output Address" button, then copy the "Zip format". For example:

(The first page address: open the second page, move mouse over link 1, copy the address of it and paste here)
(The second page address: move mouse over link 2, copy the address of it and paste here)
(Total page number: the number is shown on pages usually, copy it and paste here)
(Zip format: after clicking "Output Address" button, copy it for using later)

In InfoSeek, select "page_address" job first, right click on right-down window, select meun item "Add>>Add..." to open Add Pages dialog, then paste the "Zip format" above to the editor, click "OK" to return to the main interface. Double-click "page_address" job to start search, copy/export results (addresses of all messages) in the left window after job finish.

Step 2: Search for info of all messages (with "page_info" job of InfoSeek)

1. Customize "page_info" pattern

With "Info Pattern Maker", input scripts in "Script editor", each line is a unit, the syntax is:

* Title String<LM>	"<LM>" means the content for the title ends in the middle of the line. "Title String" denotes the title of a searched field, such as "Email, "Tel", "Fax" etc.
* Title String<LE>	"<LE>" means the content for the title ends at the end of the line.
* Title String<ML>	"<ML>" means the content for the title is multi-line.
* User-defined Title<WL>	"<WL>" means this is a line without title (the whole line is content). "User-defined Title" denotes a user-defined title for the line, it must begin with '$'.
* Anchor String	Anchor String has no corresponding search content. It denotes an end string of "<LM>" and "<ML>", beginning or end string of "<WL>".

Notice: if "Title String" before "<LM>", "<LE>" or "<ML>" doesn't exist, it can be replaced with "User-defined Title". "User-defined Title" must begin with '$'.

Example 1: Search for all fields.	Example 2: Search for fields: "Address", "Contact", "Site".	Example 3: Search for fields: "Zip", "Tel", "Overview".
Prev page $Company: <WL> Address: <LM> Zip: <LM> Contact: <LE> Tel: <LM> Fax: <LE> Email: <LM> Site: <LE> Overview: <ML> Next page	Address: <LM> Zip Contact: <LE> Site: <LE>	Zip: <LM> Contact Tel: <LM> Fax Overview: <ML> Next page

After input, click "Info Pattern" button to get the pattern and its format. Open "Options" dialog of InfoSeek (select menu "File">>"Options"), switch to "Pattern" tab, copy data to "page_info" pattern.

2. Add pages to search

In InfoSeek, select "page_info" job first, right click on right-down window, select meun item "Add>>Add..." to open Add Pages dialog, then paste results of Step 1 to its editor, click "OK" to return to the main interface. Double-click "page_info" job to start search, copy/export results in the left window after job finish.

User Interface

Address Pattern Maker:

Compare addresses backward: Search sample addresses for same tail string. If addresses have same tail string (such as "abc.chem17.com" and "derun.chem17.com"), check the item can make pattern created more precise.

Info Pattern Maker:

LM: Copy string "<LM>" to Clipboard for paste.
LE: Copy string "<LE>" to Clipboard for paste.
ML: Copy string "<ML>" to Clipboard for paste.
WL: Copy string "<WL>" to Clipboard for paste.
Minimal character number for whole line: If character number of a line is less than the value, it does not match "<WL>" (the whole line).

(Format options)
Include URL of source pages: The first item in output format is the address of source page. You can browse or check the source page with the address.

(Field attributes)
Must exist: in a result item, the field must not be null, if null, the item will not be exported.
Not output: when exporting items, the content of the field will be saved as null.
Unique: when exporting items to a database, the content of the field must not be duplicated.
Memo type: the field of memo type can own 256-32768 bytes, otherwise it is less than 256.

Analyze Format: get format options according to current content of Format and Comment, then you can adjust attributes and positions of fields.

FAQ

* Addresses created by "Batch Address Creator" have errors?
Notice how to copy the address of the first page: you should switch current page to the other page (such as the second page), then copy the link of the first page. Do not copy it in IE address box of the first page.

* In "Batch Address Creator", I have copied addresses of the first and the second page correctly, but addresses created are not able to be searched?
Check and compare different addresses, there must be one variable changes in term of number. If the number of variable is more than one or variables do not change according to number, "Batch Address Creator" is not usable here!

* Addresses as the search result of Step 1 cannot be exported, how to test step 2?
Unregistered version of InfoSeek cannot save or export results. If you wanted to test step 2, make a search result selected, select "Edit" from right-click menu, then copy the content of the result from the pop-up dialog.

* IE prompts "web page script error" or no output?
Close web page firewall of Anti-Virus softwares, make sure IE browser is used and JavaScript function has been enabled.

Tips & Tricks

* For FastSeek, you can customize targeted address patterns with "Address Pattern Maker" or infomation patterns with "Info Pattern Maker".

* When customizing info pattern, select and copy all of the text in targeted page (Ctrl+A & Ctrl+C), then analyze it (there may be some errors if analyzing the content shown in page directly).

* After customizing info pattern, copy the text of "Script editor" and save it, then you can paste it here directly next time.

* "Info Pattern Maker" is applied to customizing pages with fixed format. If you wanted to customize other complex pages, please feel free to contact us: support@allweb-soft.com.

* Please visit our site: http://www.allweb-soft.com to download the latest version of "Search Companion".

Shortcuts

Ctrl+A: Select All.
Ctrl+C: Copy.
Ctrl+V: Paste.
Ctrl+X: Cut.