Abstract: A computerized query optimizer for use with a database system having an ordered set of records. The optimizer employs a scanner and an evaluator. A query is composed as ranges of record values related by logical operators. The query is converted to a Boolean tree in canonical form. The tree is optimized to express the ranges as a set of disjoint semi-open ranges. The scanner reads a next record from the database. The evaluator, using the query, delivers a logical true or false condition for the record. In addition, the evaluator also delivers an interval of values having the same logical condition as the logical condition of the record. If this logical condition is false, the scanner skips over records having values of the interval, otherwise, if the logical condition is true, records having values of the interval are selected.

Abstract: A computerized query optimizer for use with a database system having an ordered set of records. The optimizer employs a scanner and an evaluator. A query is composed as ranges of record values related by logical operators. The query is converted to a Boolean tree in canonical form. The tree is optimized to express the ranges as a set of disjoint semi-open ranges. The scanner reads a next record from the database. The evaluator, using the query, delivers a logical true or false condition for the record. In addition, the evaluator also delivers an interval of values having the same logical condition as the logical condition of the record. If this logical condition is false, the scanner skips over records having values of the interval, otherwise, if the logical condition is true, records having values of the interval are selected.

Abstract: In a computer system, input strings to be translated are composed of characters selected from a first alphabet. According to a predetermined criterion, a list of sub-strings is selected from the input strings to form entries in a dictionary. The entries of the dictionary are arranged according to a collating order of the first alphabet. An interval including the sub-strings of the input strings is partitioned into an all-inclusive and disjoint set of ranges. The sub-strings of the interval are arranged according to the collating order of the first alphabet, and each sub-strings of a particular range has a common prefix, the common prefix selected from the list of sub-strings. A unique encoding is assigned to each common prefix, the corresponding set of unique encodings composed of characters selected from a second alphabet. The input strings are parsed, one at the time, into a plurality of tokens, each token corresponding to a sub-string selected from the dictionary.

Abstract: Dynamic optimization of a database access. During execution, based on the actual number of retrieved keys, or on the actual number of fetches (I/O operations) used to retrieve keys, a decision is made between alternative paths for completing the access. The invention features deciding not to retrieve records using the keys retrieved using a particular index if that retrieval could be expected to take longer than retrieving all records. A second index scan is begun during the first index scan, and a decision between alternative paths is made based on the number of keys retrieved during the second scan, or on the number of fetches incurred in the second scan. Scanning of an index is terminated if the number of retrieved keys exceeds a threshold corresponding to the point at which record retrieval using keys is likely to be more costly than full record retrieval. If all index scans are terminated, a full, sequential retrieval of all records is done.

Abstract: The invention is a compression method which provides the advantages of run-length compression while preserving the natural collating order of the data. The method of the invention replaces sequential repetitions of a target pattern beyond a threshold value followed by a single number which encodes both the repetition count and the collating relationship between the target pattern and the pattern which follows it in the uncollated sequence. In the preferred embodiment of the invention, the code is preceded by at least one repetition of the pattern. In the preferred embodiment of the invention, the numerical code is set as the difference between the run count and the threshold count if the following pattern is less than the target pattern, or if there is no following pattern. If the following pattern is greater than the compression sequence, the code is set as a predetermined number, preferably the maximum binary value of the sequence length, minus the difference between the run count and the threshold count.

Abstract: A random sample is obtained from an inverted tree data structure by maintaining a cardinality estimate for each intermediate node in the tree, and selecting a leaf node at random by descending from the root node and performing an acceptance/rejection selection at each intermediate node weighted proportional to the cardinality estimates of the children and weighted inversely proportional to the cardinality estimate of the intermediate node. Even though the selection of an intermediate node is based upon only an estimate of the true cardinality of the intermediate node, the error in the estimate does not cause bias in the overall sampling method because any bias in the selection of the intermediate node is cancelled by an opposite bias in the selection of the child of the intermediate node.

Abstract: In a system for compressing, decompressing, and logically manipulating arbitrary bit-maps, the bit-map is aligned along byte boundaries. Each aligned byte is classified as a gap byte if all of the bits of the byte store the same logical value, otherwise the byte is classified as a non-gap byte. Adjacent bytes of the same class are grouped. Groups of gap bytes are encoded into an atomic sequence of bytes as a count of the number of bits included in the grouped gap bytes. Map bytes are duplicated in the atomic sequence and also associated with a count thereof.