The build operation has three phases:
1. Copy phase (copying files)
2. Index phase (indexing files)
3. Write phase (write to the build destination drive) This step only occurs when you build to removable media or to a logical drive other than the drive where the you initially created the database. When you build to the hard drive where you initially created the database, the data containers stay on the hard drive, and the build process is complete once the files are copied and indexed. Alchemy never copies the search program to the build destination in this situation.
When you build a database, all newly added files and custom profile data are copied to data container files, which are described below:
data000 folder: This subfolder is created under the Alchemy folder, and contains the following files:
00001.dat: Document files are copied to this file. This file is created if you choose to copy files to data containers when you build a database. By default, uncompressed files are compressed, but already compressed files are not compressed further. All original links that may exist in HTML or PDF documents are retained in this file. A new data file is created after the current file reaches the specified buffer size, and automatically increments the number. (The second data file is named 00002.dat, and so on.)
00001.pro: Custom profile data is copied to this file. System profile information is stored in the database TOB file. The same incremental naming scheme described for the data file applies to the profile file. Each data file has a corresponding profile file; one pair is created for each buffer-sized group of files. Each data000 folder holds up to 100 data files and 100 corresponding profile files. After the limit is reached, a second data folder is created and increments the number in the folder name to data001. Data folders are created in this way up to 999 data folders, each one containing 100 sets of data and profile files. Alchemy continues to copy files to data containers until the container file is full or until the last file is copied. Then Alchemy indexes the data in the container files.
Once the container files are full or the last file is copied, then Alchemy opens and indexes each file in the DAT file. The indexed information is temporarily stored in the database_name.tdx file (in an uncompressed format) in the alw_work folder in your system’s Temp folder. Then Alchemy compresses the index files and merges them with the current NDX file if it exists from a previous build. The existing database_name.ndx file is renamed to the database_name.tdx, and after the entire process is complete it is renamed to database_name.ndx. When building to the local hard drive, the build process is complete after files in the data containers are in-dexed. When building to removable media or a network drive, after indexing the files in the data containers, Alchemy writes the data container and index files to the build destination.
This phase occurs when you build to removable media or to a logical drive other than the drive where you initially created the database because Alchemy needs to write the information contained in the database to the new location at the time of the build. During the writing phase, Alchemy copies data container files, index files, and the search program to the build destination. After the files are successfully copied, the data container files are deleted from the hard drive.
CAUTIONS: When you build to removable media or to a logical drive other than the drive where you initially created the database, you should set Alchemy to copy the files and the profile information to data containers. The information for the source files will not exist at the new location unless you copy the file information into data containers. When Alchemy copies this information to a data container, the container either reaches the designated buffer size, in which case Alchemy creates a new container for the remaining information, or Alchemy copies the last file to the container. Next, Alchemy indexes the data in the container(s) and merges newly indexed data with previously indexed data. Finally, Alchemy copies the data container files, indexes, and search pro-gram to the build destination.