JP4034591B2

JP4034591B2 - Structured document storage method and retrieval method thereof

Info

Publication number: JP4034591B2
Application number: JP2002133065A
Authority: JP
Inventors: 吉孝中村; 雅之寺田
Original assignee: Nippon Telegraph and Telephone Corp
Current assignee: Nippon Telegraph and Telephone Corp
Priority date: 2002-05-08
Filing date: 2002-05-08
Publication date: 2008-01-16
Anticipated expiration: 2022-05-08
Also published as: JP2003330938A

Description

【０００１】
【発明の属する技術分野】
本発明は、構造化文書を格納、検索する方法に関し、特に一意の文書識別情報が定義でき、参照による継承関係が定義された構造化文書を格納し、効率良く検索する方法に関するものである。
【０００２】
【従来の技術】
継承関係が定義された構造化文書の管理方法に関しては、電子的な権利情報、価値情報を記述、管理する方法として特開２００２−４１７００が提案されている。この公知例では、構造化文書等で実現される権利定義を単なる構造化文書としてではなく、権利内容を構成する各種の属性を定義するためのデータ構造として捉え、さらにオブジェクト指向の考え方である継承の概念を取り入れることで、上位概念を定義した構造化文書を継承してより具体化した権利内容を定義することを可能にし、類似の権利定義を多量に作成・管理することを容易にしている。
【０００３】
しかし、この公知例では多量の権利定義を効率良く蓄積、取得する方法について記載されていない。
【０００４】
一方、ＳＧＭＬ文書やＸＭＬ文書を含む構造化文書を格納・検索するシステムとしては、これまで多くの提案がなされており、例えば特開２０００−９００９１、特開２００１−３４６１８、特開２００１−１３４５９６等が知られている。
【０００５】
しかし、これらの構造化文書の格納・検索方法は、ある構造化文書や、そこからハイパーリンクで連結された複数の構造化文書の集合を、構造化文書としての構文を予め解析し、論理構造を内部データ化して格納するものであり、継承関係の定義された構造化文書の格納・検索に適用するには以下のような問題点があった。
【０００６】
即ち、新規に構造化文書を格納する際、該構造化文書全体の構文解析を行う必要があるため、処理コストが高くなったり、論理構造化してしまうと失われる個々の構造化文書の継承関係や個別の記述内容を用いた検索が困難になったりするという問題である。
【０００７】
【発明が解決しようとする課題】
前記の様に、従来の構造化文書の格納・検索方法では、継承関係が定義された構造化文書を効率良く格納し、検索することができなかった。
【０００８】
本発明はこの点に鑑みてなされたもので、その目的は、継承付き構造化文書において、テーブル構成と格納方法、検索方法を工夫することで、格納時に毎回全ての継承構造化文書について内容を辿って論理構造を生成しなくても、継承関係が定義された構造化文書の効率の良い格納・検索を可能にすることである。
【０００９】
【課題を解決するための手段】
前記目的を達成するため、本発明では、継承親となる構造化文書に対応する文書識別情報を格納する親情報カラムと、継承子となる構造化文書に対応する文書識別情報を格納する子情報カラムとからなるレコードを少なくとも１つ含む継承関係テーブルを用いることで、構造化文書の内容を解釈したり、論理構造化したりすることなく、継承関係を辿ることが可能となる。
【００１０】
また、本発明では、継承関係テーブルの各レコードに、親情報カラムの文書識別情報に対応する構造化文書と子情報カラムの文書識別情報に対応する構造化文書との相対的な継承段数を格納する相対段数カラムを設けることで、ある構造化文書が継承する複数の構造化文書について構造化文書の内容を解析したり、継承関係テーブルへの複数回の参照をせずに論理的な継承関係を取得でき、構造化文書の論理構成を容易に再構成できる。
【００１１】
また、本発明では、前記継承関係テーブルとともに、文書識別情報を格納する識別情報カラムと、予め指定された内容情報の一部分である検索対象領域についてその領域に記述された値を格納する少なくとも１つの検索対象領域カラムとからなるレコードを少なくとも１つ含む検索インデックステーブルを用いることで、論理構造を生成することなく、構造化文書内の値による部分一致検索が可能となる。
【００１２】
また、本発明では、検索インデックステーブルの各レコードに、各検索対象領域カラム毎に該カラム内の値が当該構造化文書の内容部で初めて記述されている場合は真値を取り、それ以外の場合は偽値を取る出現フラグを格納する出現フラグカラムを設けることで、論理構造の再構成後には分からなくなってしまう、当該検索対象領域を有する構造化文書の物理的実体の特定を可能とする。
【００１３】
【発明の実施の形態】
以下、図面に従って本発明の実施の形態について説明する。
【００１４】
図１は本発明で格納・検索対象とする構造化文書の基本構成を示すもので、当該構造化文書１０は参照先（親）の構造化文書を指定する情報として個々の構造化文書に対応する一意の文書識別情報を零（参照先無し）または一つのみ含む参照情報部１１と、当該構造化文書１０の内容情報を含む内容部１２とからなっている。
【００１５】
ここで、参照元の構造化文書と該参照元の構造化文書の参照情報部に文書識別情報で指定された参照先の構造化文書との間に定義される継承関係とは、参照元の構造化文書の論理構造が参照先の構造化文書の論理構造を包含し、参照先の構造化文書の論理構造の任意の一部分への参照処理で得られる結果と、参照元の構造化文書の論理構造に参照処理をして得られる結果とが一致するという意味で「参照元の構造化文書ｉｓａ参照先の構造化文書」と解釈される関係（ｉｓ−ａ関係）である。
【００１６】
なお、構造化文書の論理構造とは、該構造化文書を出発点として、参照情報部に参照先の構造化文書の文書識別情報が記述されていない構造化文書まで遡っていった場合の全ての構造化文書の内容情報をつなげたものである。
【００１７】
図２は構造化文書の継承関係の一例を示すもので、ここでは６個の構造化文書１０ａ，１０ｂ，１０ｃ，１０ｄ，１０ｅ，１０ｆにおいて、構造化文書１０ｂ，１０ｃの参照情報部に構造化文書１０ａの文書識別情報が記述され、構造化文書１０ｄの参照情報部に構造化文書１０ｂの文書識別情報が記述され、構造化文書１０ｅ，１０ｆの参照情報部に構造化文書１０ｃの文書識別情報が記述されている例（構造化文書１０ａの参照情報部には記述無し）を示す。
【００１８】
図３は本発明で格納・検索対象とする構造化文書の具体例を示すもので、ここでは特開２００２−４１７００に記載された権利定義を想定した例、詳細には以下の５つのタグセットと１つの属性値を有するＸＭＬ文書の集合である権利定義２０の例を示す。
【００１９】
Ｔｉｃｋｅｔタグ：権利定義のトップエレメントであり、Ｐａｒｅｎｔ属性を有することができる。
【００２０】
Ｔｙｐｅタグ：チケットの種別を記述するエレメント。本実施の形態では値「コンサート」を有する権利定義の集合を例に採る。
【００２１】
Ｎａｍｅタグ：チケットの名称を記述するエレメント。本実施の形態では値「クリスマスコンサート」を有する権利定義の集合を例に採る。
【００２２】
Ｐｌａｃｅタグ：場所を記述するエレメント。本実施の形態では値「東京」「名古屋」「大阪」を有する権利定義の集合を例に採る。
【００２３】
Ｇａｔｅタグ：チケットの改札場所を記述するエレメント。本実施の形態では値「Ｇａｔｅ１」「Ｇａｔｅ２」「Ｇａｔｅ３」を有する権利定義の集合を例に採る。
【００２４】
Ｐａｒｅｎｔ属性：トップエレメントで指定され、継承する権利定義（親権利定義）の文書識別情報である権利定義ＩＤを指定する。
【００２５】
ここで、Ｔｉｃｋｅｔタグ内のＰａｒｅｎｔ属性が前述した参照情報部に該当し、それ以外が前述した内容部に該当する。
【００２６】
図４は前述した権利定義の具体例、ここでは９個の権利定義２０Ａ，２０Ｂ，２０Ｃ，２０Ｄ，２０Ｅ，２０Ｆ，２０Ｇ，２０Ｈ，２０Ｉを、また、図５は図４に示した権利定義２０Ａ〜２０Ｉの継承関係を表すツリーを示すものである。
【００２７】
なお、権利定義の文書識別情報である権利定義ＩＤは当該権利定義のハッシュ値をとることとし、本実施の形態ではＡ，Ｂ，Ｃ，Ｄ，Ｅ，Ｆ，Ｇ，Ｈ，Ｉ等のアルファベットでその値を表すこととする。
【００２８】
ここで、ハッシュ値とは、ｙ＝Ｈ（ｘ）であるようなｘをｙから求めることが極めて困難な一方向関数であり、かつ異なる入力から得られる結果が一致する確率が極めて低いという性質を有するハッシュ関数により入力のデータから得られる値である。この様なハッシュ関数としてはＭＤ５やＳＨＡ−１などが広く知られている。
【００２９】
本実施の形態では、権利定義の参照情報部をトップエレメントの属性として表したが、例えばタグの一つとして記述することも可能である。また、文書識別情報として親権利定義のハッシュ値である権利定義ＩＤとしたが、権利定義をファイルとして管理する場合にはそのＵＲＬを指定することも可能である。
【００３０】
図６は本発明方法を実現する装置構成の一例を示すもので、リレーショナルデーベース（ＲＤＢＳ）内に格納された３つのテーブル、即ちメインテーブル３０、継承関係テーブル（親子テーブル）４０及び検索インデックステーブル５０と、文書格納処理部６０と、文書検索処理部７０とから構成される。
【００３１】
メインテーブル３０は、図７に示すように、構造化文書を格納する文書カラム（権利定義カラム）３１と、該構造化文書に対応する文書識別情報を格納する文書識別情報カラム（権利定義ＩＤカラム）３２とからなるレコードを少なくとも１つ含む。
【００３２】
継承関係テーブル４０は、図８に示すように、継承親となる構造化文書に対応する文書識別情報を格納する親情報カラム（親権利定義ＩＤカラム）４１と、継承子となる構造化文書に対応する文書識別情報を格納する子情報カラム（子権利定義ＩＤカラム）４２と、親情報カラムの文書識別情報に対応する構造化文書と子情報カラムの文書識別情報に対応する構造化文書との相対的な継承段数を格納する相対段数カラム４３とからなるレコードを少なくとも１つ含む。
【００３３】
検索インデックステーブル５０は、図９に示すように、文書識別情報を格納する識別情報カラム（権利定義ＩＤカラム）５１と、予め指定された内容情報の一部分である検索対象領域についてその領域に記述された値を格納する少なくとも１つの検索対象領域カラム（Ｔｙｐｅカラム、Ｎａｍｅカラム、Ｐｌａｃｅカラム、Ｇａｔｅカラム）５２と、各検索対象領域カラム毎に該カラム内の値が当該構造化文書の内容部に記述されている場合は真値（ＴＲＵＥ）を取り、それ以外の場合は偽値（ＦＡＬＳＥ）を取る出現フラグを格納する出現フラグカラム（Ｔｙｐｅフラグカラム、Ｎａｍｅフラグカラム、Ｐｌａｃｅフラグカラム、Ｇａｔｅフラグカラム）５３とからなるレコードを少なくとも１つ含む。
【００３４】
文書格納処理部６０及び文書検索処理部７０は、前記ＲＤＢＳにアクセス可能なコンピュータとその上で実行される文書格納プログラム及び文書検索プログラムによって実現される。
【００３５】
図１０乃至図１９は文書格納プログラム及び文書検索プログラムに対応する処理の流れ図である。詳細には、図１０は請求項１の文書格納方法、図１１は請求項２の文書格納方法、図１２は請求項４の文書格納方法、図１３は請求項６の文書格納方法にそれぞれ対応する文書格納プログラムの流れ図である。また、図１４は請求項３の文書検索方法、図１５は請求項５の文書検索方法、図１６，図１７，図１８，図１９は請求項７の文書検索方法にそれぞれ対応する文書検索プログラムの流れ図である。
【００３６】
なお、図１５において「検索対象領域のうちのいくつかとその検索対象領域に記述された値との組によるＡＮＤ条件」は「ＡＮＤ条件」と、「検索対象領域のうちのいくつかとその検索対象領域に記述された値との組と文書識別情報とによるＡＮＤ条件」は「文書識別情報を含むＡＮＤ条件」と略記した。また、図１６において「検索対象領域のうちのいくつかとその検索対象領域に記述された値との組によるＡＮＤ条件並びに子排除フラグ」は「ＡＮＤ条件と子排除フラグ」と、「検索対象領域のうちのいくつかとその検索対象領域に記述された値との組と文書識別情報とによるＡＮＤ条件並びに子排除フラグ」は「文書識別情報を含むＡＮＤ条件と子排除フラグ」とそれぞれ略記した。また、図１６中の「別紙フローα」、「別紙フローβ」、「別紙フローγ」はそれぞれ図１７、図１８、図１９を示す。
【００３７】
以下、権利定義２０Ａ，２０Ｂ，２０Ｃが格納されている状態から、権利定義２０Ｄを新たに格納する場合の処理手順を図２０を用いて説明する。
【００３８】
（１）権利定義２０Ｄのハッシュ値として、権利定義ＩＤ＝Ｄを作成（ｓ１）。
【００３９】
（２）メインテーブル３０に権利定義ＩＤと権利定義を格納（ｓ２）。
【００４０】
（３）権利定義２０ＤからＰａｒｅｎｔ属性の値Ｃを取得（ｓ７）。
【００４１】
（４）継承関係テーブル４０から、子カラムにＣを有するレコードを取得（ｓ４）。
【００４２】
集合１：｛親，子，段数｝＝｛Ａ，Ｃ，２｝，｛Ｂ，Ｃ，１｝，｛Ｃ，Ｃ，０｝が取得される。
【００４３】
（５）集合１の３つのレコードの親カラムの値Ａ，Ｂ，Ｃを親カラムに有し、Ｄを子カラムに有し、相対段数にそれぞれ１を加えた値３，２，１を相対段数カラムに有するレコードと、親カラム、子カラムともＤを有し、相対段数カラムに０を有するレコードを継承関係テーブル４０に追加（ｓ５’，ｓ６’）。
【００４４】
レコード｛親，子，段数｝＝｛Ａ，Ｄ，３｝，｛Ｂ，Ｄ，２｝，｛Ｃ，Ｄ，１｝，｛Ｄ，Ｄ，０｝が追加される。
【００４５】
（６）検索インデックステーブル５０から、権利定義ＩＤカラムにＣを有するレコードを取得（ｓ８）。
【００４６】
（７）検索インデックステーブル５０に、権利定義ＩＤカラムに値Ｄを有し、Ｃのレコードの各検索対象領域カラムの値を当該カラムに有し、対応するフラグカラムにはＦＡＬＳＥを有し、Ｃのレコードに記載された以外の検索対象領域であるＧａｔｅの値Ｇａｔｅ１を取得してＤのレコードのＧａｔｅカラムに格納し、Ｇａｔｅフラグに値ＴＲＵＥを格納するレコードを追加（ｓ９’）。
【００４７】
また、以下に、いくつかの検索条件と、検索処理の概要、取得される権利定義ＩＤを例示する。
【００４８】
検索例１：「権利定義ＩＤ＝Ｆ」を条件とする検索
継承関係テーブル４０から、「親権利定義ＩＤ＝Ｆ」のレコードの子権利定義ＩＤ｛Ｆ，Ｉ｝を取得し、これを検索結果とする（ｓ１１〜ｓ１３）。
【００４９】
子排除フラグにＴＲＵＥが指定されていた場合、「親権利定義ＩＤ＝子権利定義ＩＤ＝Ｆ」のレコードの存在を確認し、存在するので｛Ｆ｝を検索結果とする（ｓ１４，ｓ２１〜ｓ２４）。
【００５０】
検索例２：「＜Ｇａｔｅ＞＝Ｇａｔｅ１」を条件とする検索
検索インデックステーブル５０から、「＜Ｇａｔｅ＞＝Ｇａｔｅ１」のレコードの権利定義ＩＤ｛Ｄ，Ｉ｝を取得し、これを検索結果とする（ｓ１１，ｓ１５，１６）。
【００５１】
子排除フラグにＴＲＵＥが指定されていた場合、この結果のうちＧａｔｅフラグがＴＲＵＥである｛Ｄ，Ｉ｝を検索結果とする（ｓ１４，ｓ３１〜ｓ３３）。
【００５２】
検索例３：「＜Ｎａｍｅ＞＝クリスマスコンサートＡＮＤ＜Ｐｌａｃｅ＞＝東京」を条件とする検索
検索インデックステーブル５０から、「＜Ｎａｍｅ＞＝クリスマスコンサート」かつ「＜Ｐｌａｃｅ＞＝東京」のレコードの権利定義ＩＤ｛Ｃ，Ｄ，Ｇ，Ｈ｝を取得し、これを検索結果とする（ｓ１１，ｓ１４〜１６）。
【００５３】
子排除フラグにＴＲＵＥが指定されていた場合、この結果のうち、ＮａｍｅフラグまたはＰｌａｃｅフラグがＴＲＵＥである｛Ｃ｝を検索結果とする（ｓ３１〜ｓ３３）。
【００５４】
検索例４：「＜Ｎａｍｅ＞＝クリスマスコンサートＡＮＤＩＤ＝Ｃ」を条件とする検索
継承関係テーブル４０から、「親権利定義ＩＤ＝Ｃ」の子権利定義ＩＤ｛Ｃ，Ｄ，Ｇ，Ｈ｝を取得し（ｓ１１，ｓ１４，ｓ１７）、検索インデックステーブル５０から、これらを権利定義ＩＤに有するレコードの集合を取得する（ｓ１８）。該レコードの集合から、「＜Ｎａｍｅ＞＝クリスマスコンサート」のレコードの権利定義ＩＤ｛Ｃ，Ｄ，Ｇ，Ｈ｝を取得し（＊）、これを検索結果とする（ｓ１９）。
【００５５】
子排除フラグにＴＲＵＥが指定されていた場合、この結果のうち、ＮａｍｅフラグがＴＲＵＥなものを検索するが、該当がないため、｛Ｃ｝を検索結果とする（ｓ４１〜ｓ４７）（＊の段階で既に「親権利定義ＩＤ＝子権利定義ＩＤ＝Ｃ」の存在が確認されているため）。
【００５６】
また、前記検索例で得られた結果である権利定義ＩＤの各々について、その論理構造を継承段数を用いて再構成する処理手順を説明する。
【００５７】
例えば、権利定義ＩＤ＝Ｈの論理構造を取得するためには、
（１）継承関係テーブル４０から、「子権利定義ＩＤ＝Ｈ」のレコードを取得し、その親権利定義ＩＤを継承段数の降順に並べて継承関係を再構成する。
【００５８】
この場合、｛親権利定義ＩＤ，継承段数｝＝｛Ａ，３｝｛Ｂ，２｝｛Ｃ，１｝｛Ｈ，０｝なので、継承親から順に、Ａ−Ｂ−Ｃ−Ｈという継承関係が得られる。
【００５９】
（２）それぞれの権利定義ＩＤについて、メインテーブル３０から権利定義を取得し、その内容部を先頭から順に連結することで論理構造を得る。
【００６０】
即ち、権利定義ＩＤ＝Ｈの権利定義は、論理構造として以下の内容情報と同一視される。
【００６１】
＜Ｔｙｐｅ＞コンサート＜／Ｔｙｐｅ＞
＜Ｎａｍｅ＞クリスマスコンサート＜／Ｎａｍｅ＞
＜Ｐｌａｃｅ＞東京＜／Ｐｌａｃｅ＞
＜Ｇａｔｅ＞Ｇａｔｅ３＜／Ｇａｔｅ＞
【００６２】
【発明の効果】
以上説明したように、請求項１及び３の発明によれば、継承関係テーブルを用いることで、構造化文書の参照情報部から参照先を取得して逐次に継承関係を辿ることなく、ある構造化文書が継承する構造化文書の集合を特定することができる。また、予め継承関係をツリー構造化して記憶しておかなくても、単純なテーブル参照によりある構造化文書を継承している全ての子構造化文書を特定することができる。
【００６３】
また、請求項２及び３の発明によれば、継承段数を管理することで、ある構造化文書が継承する複数の構造化文書について、参照情報部を逐次に解析したり、継承関係テーブルへの参照を逐次に行ったりせずに、ある文書識別情報を有する構造化文書の継承関係が定義された全ての構造化文書の継承階層構造を取得でき、当該構造化文書の論理構造、即ち自らを含む継承関係が定義された全ての構造化文書の内容部を継承階層順に連結した文書を容易に再構成することができる。
【００６４】
また、請求項４及び５の発明によれば、構造化文書内の特定の検索対象領域に対するインデックスを張った検索インデックステーブルを用いることで、全ての構造化文書の論理構造を予め取得しなくても、論理構造に対する部分一致検索を行うことができる。
【００６５】
また、請求項６及び７の発明によれば、検索インデックステーブルに、検索対象領域の出現フラグを設けることで、論理構造の再構成後には情報が失われてしまう当該検索対象領域を有する構造化文書の物理的実体の特定が可能となる。
【図面の簡単な説明】
【図１】本発明で対象とする構造化文書の基本構成を示す模式図
【図２】構造化文書の継承関係の一例を示す模式図
【図３】本発明で対象とする構造化文書の具体例を示す模式図
【図４】権利定義の具体例を示す模式図
【図５】図４に示した権利定義の継承ツリーの模式図
【図６】本発明方法を実現する装置構成の一例を示すブロック図
【図７】図６中のメインテーブルの一例を示す模式図
【図８】図６中の継承関係テーブルの一例を示す模式図
【図９】図６中の検索インデックステーブルの一例を示す模式図
【図１０】請求項１の文書格納方法に対応する文書格納プログラムの流れ図
【図１１】請求項２の文書格納方法に対応する文書格納プログラムの流れ図
【図１２】請求項４の文書格納方法に対応する文書格納プログラムの流れ図
【図１３】請求項６の文書格納方法に対応する文書格納プログラムの流れ図
【図１４】請求項３の文書検索方法に対応する文書格納プログラムの流れ図
【図１５】請求項５の文書検索方法に対応する文書格納プログラムの流れ図
【図１６】請求項７の文書検索方法に対応する文書格納プログラムの流れ図
【図１７】請求項７の文書検索方法に対応する文書格納プログラムの流れ図
【図１８】請求項７の文書検索方法に対応する文書格納プログラムの流れ図
【図１９】請求項７の文書検索方法に対応する文書格納プログラムの流れ図
【図２０】文書格納処理の一例を示す説明図
【符号の説明】
３０：メインテーブル、３１：文書カラム（権利定義カラム）、３２：文書識別情報カラム（権利定義ＩＤカラム）、４０：継承関係テーブル、４１：親情報カラム（親権利定義ＩＤカラム）、４２：子情報カラム（子権利定義ＩＤカラム）、４３：相対段数カラム、５０：検索インデックステーブル、５１：識別情報カラム（権利定義ＩＤカラム）、５２：検索対象領域カラム（Ｔｙｐｅカラム、Ｎａｍｅカラム、Ｐｌａｃｅカラム、Ｇａｔｅカラム）、５３：出現フラグカラム（Ｔｙｐｅフラグカラム、Ｎａｍｅフラグカラム、Ｐｌａｃｅフラグカラム、Ｇａｔｅフラグカラム）、６０：文書格納処理部、７０：文書検索処理部。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a method for storing and retrieving structured documents, and more particularly to a method for storing and retrieving structured documents in which unique document identification information can be defined and inheritance relationships by reference are defined.
[0002]
[Prior art]
As a method for managing structured documents in which inheritance relationships are defined, Japanese Patent Laid-Open No. 2002-41700 has been proposed as a method for describing and managing electronic rights information and value information. In this well-known example, the rights definition realized in a structured document etc. is not just a structured document, but as a data structure for defining various attributes constituting the content of rights, and inheritance is an object-oriented concept. By adopting the concept, it is possible to define more specific rights contents by inheriting the structured document that defines the superordinate concept, making it easy to create and manage a large number of similar rights definitions .
[0003]
However, this known example does not describe a method for efficiently accumulating and acquiring a large amount of right definitions.
[0004]
On the other hand, as a system for storing / retrieving structured documents including SGML documents and XML documents, many proposals have been made so far. For example, JP 2000-90091, JP 2001-34618, JP 2001-134596, etc. It has been known.
[0005]
However, these structured document storage / retrieval methods analyze the syntax of a structured document or a set of structured documents linked by hyperlinks from the structured document in advance as a logical structure. Are stored as internal data, and there are the following problems when applied to storage / retrieval of structured documents in which inheritance relationships are defined.
[0006]
That is, when a new structured document is stored, it is necessary to parse the entire structured document, so that the inheritance relationship of individual structured documents that is lost when the processing cost increases or logically structured Or the search using individual description contents becomes difficult.
[0007]
[Problems to be solved by the invention]
As described above, the conventional structured document storage / retrieval method cannot efficiently store and retrieve structured documents in which inheritance relationships are defined.
[0008]
The present invention has been made in view of this point, and its purpose is to devise the table configuration, storage method, and search method in a structured document with inheritance, so that the contents of all inherited structured documents are stored every time when stored. Even without tracing and generating a logical structure, it is possible to efficiently store and search a structured document in which an inheritance relationship is defined.
[0009]
[Means for Solving the Problems]
In order to achieve the above object, in the present invention, a parent information column that stores document identification information corresponding to a structured document that becomes an inheritance parent, and child information that stores document identification information corresponding to a structured document that becomes an inheritance child By using an inheritance relationship table including at least one record consisting of columns, it is possible to trace the inheritance relationship without interpreting the contents of the structured document or logically structuring it.
[0010]
In the present invention, the relative number of inheritance stages between the structured document corresponding to the document identification information in the parent information column and the structured document corresponding to the document identification information in the child information column is stored in each record of the inheritance relationship table. By providing a relative column number column, the contents of a structured document can be analyzed for multiple structured documents inherited by a structured document, and the logical inheritance relationship can be obtained without multiple references to the inheritance relationship table. And the logical structure of the structured document can be easily reconfigured.
[0011]
In the present invention, together with the inheritance relation table, an identification information column for storing document identification information, and at least one value for storing a value described in the area for a search target area that is a part of content information designated in advance. By using a search index table including at least one record including a search target area column, partial match search using values in the structured document can be performed without generating a logical structure.
[0012]
In the present invention, each record in the search index table takes a true value when the value in the column is described for the first time in the content part of the structured document for each search target area column. In this case, by providing an appearance flag column for storing an appearance flag that takes a false value, it becomes possible to identify the physical entity of the structured document having the search target area, which will be unknown after the logical structure is reconstructed. .
[0013]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings.
[0014]
FIG. 1 shows a basic structure of a structured document to be stored / retrieved in the present invention. The structured document 10 corresponds to each structured document as information for designating a reference destination (parent) structured document. The reference information part 11 includes zero or no unique document identification information to be performed, and the content part 12 includes the content information of the structured document 10.
[0015]
Here, the inheritance relationship defined between the reference source structured document and the reference destination structured document specified by the document identification information in the reference information section of the reference source structured document is the reference source The logical structure of the structured document includes the logical structure of the referenced structured document. The result obtained by referring to any part of the logical structure of the referenced structured document and the structure of the referenced structured document It is a relationship (is-a relationship) that is interpreted as “referenced structured document is a referenced destination structured document” in the sense that the result obtained by performing the reference processing on the logical structure matches.
[0016]
Note that the logical structure of a structured document refers to all cases where the structured document is traced back to a structured document in which the document identification information of the referenced structured document is not described in the reference information section. The content information of the structured documents of the above is connected.
[0017]
FIG. 2 shows an example of inheritance relationships of structured documents. Here, six structured documents 10a, 10b, 10c, 10d, 10e, and 10f are structured in the reference information section of the structured documents 10b and 10c. Document identification information of the document 10a is described, document identification information of the structured document 10b is described in the reference information portion of the structured document 10d, and document identification information of the structured document 10c is described in the reference information portions of the structured documents 10e and 10f. Is described (no description in the reference information part of the structured document 10a).
[0018]
FIG. 3 shows a specific example of a structured document to be stored / retrieved in the present invention. Here, an example assuming a rights definition described in Japanese Patent Laid-Open No. 2002-41700, specifically, the following five tag sets: And an example of a right definition 20 that is a set of XML documents having one attribute value.
[0019]
Ticket tag: A top element of a right definition, and can have a Parent attribute.
[0020]
Type tag: An element describing the type of ticket. In this embodiment, a set of right definitions having the value “concert” is taken as an example.
[0021]
Name tag: An element describing the name of a ticket. In the present embodiment, a set of right definitions having the value “Christmas concert” is taken as an example.
[0022]
Place tag: An element describing a place. In the present embodiment, a set of right definitions having values “Tokyo”, “Nagoya”, and “Osaka” is taken as an example.
[0023]
Gate tag: An element describing the ticket gate location. In the present embodiment, a set of right definitions having values “Gate 1”, “Gate 2”, and “Gate 3” is taken as an example.
[0024]
Parent attribute: Designates a right definition ID, which is specified in the top element and is document identification information of the inherited right definition (parent right definition).
[0025]
Here, the Parent attribute in the Ticket tag corresponds to the reference information part described above, and the other corresponds to the content part described above.
[0026]
FIG. 4 shows a specific example of the right definition described above, here, nine right definitions 20A, 20B, 20C, 20D, 20E, 20F, 20G, 20H and 20I, and FIG. 5 shows the right definition 20A shown in FIG. It shows a tree representing the inheritance relationship of ~ 20I.
[0027]
The right definition ID, which is the document identification information of the right definition, takes a hash value of the right definition. In this embodiment, alphabets such as A, B, C, D, E, F, G, H, and I are used. The value is represented by.
[0028]
Here, the hash value is a one-way function in which it is extremely difficult to obtain x from y such that y = H (x), and the probability that the results obtained from different inputs match is extremely low. A value obtained from input data by a hash function having As such a hash function, MD5, SHA-1, and the like are widely known.
[0029]
In the present embodiment, the reference information part of the right definition is represented as the attribute of the top element, but it can also be described as one of the tags, for example. Further, although the right definition ID, which is the hash value of the parent right definition, is used as the document identification information, when managing the right definition as a file, it is also possible to specify the URL.
[0030]
FIG. 6 shows an example of an apparatus configuration for realizing the method of the present invention. Three tables stored in a relational database (RDBS), that is, a main table 30, an inheritance relationship table (parent-child table) 40, and a search index table. 50, a document storage processing unit 60, and a document search processing unit 70.
[0031]
As shown in FIG. 7, the main table 30 includes a document column (right definition column) 31 that stores a structured document, and a document identification information column (right definition ID column) that stores document identification information corresponding to the structured document. ) 32 at least one record.
[0032]
As shown in FIG. 8, the inheritance relationship table 40 includes a parent information column (parent right definition ID column) 41 that stores document identification information corresponding to a structured document that is an inheritance parent, and a structured document that is an inheritance child. A child information column (child right definition ID column) 42 for storing corresponding document identification information, a structured document corresponding to the document identification information in the parent information column, and a structured document corresponding to the document identification information in the child information column It includes at least one record including a relative column number column 43 that stores a relative inherited column number.
[0033]
In the search index table 50, as shown in FIG. 9, an identification information column (right definition ID column) 51 for storing document identification information and a search target area which is a part of content information designated in advance are described in that area. At least one search target region column (Type column, Name column, Place column, Gate column) 52 for storing the value and the value in the column for each search target region column is described in the content part of the structured document. Appears flag columns (true flag column, name flag column, place flag column, and gate flag column) that store an appearance flag that takes a true value (TRUE) if it has been set and takes a false value (FALSE) otherwise 53 is included.
[0034]
The document storage processing unit 60 and the document search processing unit 70 are realized by a computer that can access the RDBS and a document storage program and a document search program that are executed on the computer.
[0035]
10 to 19 are flowcharts of processes corresponding to the document storage program and the document search program. Specifically, FIG. 10 corresponds to the document storage method of claim 1, FIG. 11 corresponds to the document storage method of claim 2, FIG. 12 corresponds to the document storage method of claim 4, and FIG. 13 corresponds to the document storage method of claim 6. 3 is a flowchart of a document storage program to be executed. 14 is a document search method according to claim 3, FIG. 15 is a document search method according to claim 5, and FIGS. 16, 17, 18, and 19 are document search programs corresponding to the document search method according to claim 7, respectively. It is a flowchart.
[0036]
In FIG. 15, “AND conditions based on combinations of some of the search target areas and values described in the search target areas” are “AND conditions” and “some of the search target areas and their search target areas. "AND condition based on a set of values described in the above and document identification information" is abbreviated as "AND condition including document identification information". Further, in FIG. 16, “AND condition and child exclusion flag based on a combination of some of the search target areas and values described in the search target area” are “AND condition and child exclusion flag” and “search target area “AND conditions and child exclusion flags based on a set of some of them and values described in the search target area and document identification information and child exclusion flags” are abbreviated as “AND conditions including document identification information and child exclusion flags”, respectively. Further, “attached sheet flow α”, “attached sheet flow β”, and “attached sheet flow γ” in FIG. 16 indicate FIGS. 17, 18, and 19, respectively.
[0037]
Hereinafter, a processing procedure when a right definition 20D is newly stored from a state in which the right definitions 20A, 20B, and 20C are stored will be described with reference to FIG.
[0038]
(1) Create a right definition ID = D as a hash value of the right definition 20D (s1).
[0039]
(2) The right definition ID and right definition are stored in the main table 30 (s2).
[0040]
(3) The value C of the Parent attribute is acquired from the right definition 20D (s7).
[0041]
(4) A record having C in the child column is acquired from the inheritance relationship table 40 (s4).
[0042]
Set 1: {parent, child, number of stages} = {A, C, 2}, {B, C, 1}, {C, C, 0} are acquired.
[0043]
(5) The parent column values A, B, and C of the three records in set 1 are included in the parent column, D is the child column, and the relative values are 3, 3, and 1, each of which is obtained by adding 1 to the relative number of stages. A record having a column number column, a parent column and a child column having D, and a record having a relative column number column of 0 are added to the inheritance relationship table 40 (s5 ′, s6 ′).
[0044]
Records {parent, child, number of steps} = {A, D, 3}, {B, D, 2}, {C, D, 1}, {D, D, 0} are added.
[0045]
(6) A record having C in the right definition ID column is acquired from the search index table 50 (s8).
[0046]
(7) The search index table 50 has the value D in the right definition ID column, the value of each search target area column of the record of C in the column, the corresponding flag column has FALSE, and C A Gate value Gate1 which is a search target area other than those described in the record of No. is acquired and stored in the Gate column of the record of D, and a record storing the value TRUE is added to the Gate flag (s9 ′).
[0047]
Also, some search conditions, an outline of the search process, and the acquired right definition ID will be exemplified below.
[0048]
Search example 1: The child right definition ID {F, I} of the record of “parent right definition ID = F” is acquired from the search inheritance relationship table 40 with “right definition ID = F” as a condition, and this is used as a search result. (S11 to s13).
[0049]
If TRUE is specified for the child exclusion flag, the existence of the record of “parent right definition ID = child right definition ID = F” is confirmed, and since it exists, {F} is taken as the search result (s14, s21 to s24) ).
[0050]
Search example 2: The right definition ID {D, I} of the record of “<Gate> = Gate1” is acquired from the search search index table 50 with “<Gate> = Gate1” as a condition, and this is used as the search result. (S11, s15, 16).
[0051]
If TRUE is specified for the child exclusion flag, {D, I} of which the Gate flag is TRUE is taken as the search result (s14, s31 to s33).
[0052]
Search example 3: “<Name> = Christmas concert AND <Place> = Tokyo” is used as a condition for the search definition index table 50 to define the rights of the records “<Name> = Christmas concert” and “<Place> = Tokyo”. ID {C, D, G, H} is acquired and used as a search result (s11, s14-16).
[0053]
If TRUE is specified as the child exclusion flag, {C} of which the Name flag or the Place flag is TRUE is set as the search result (s31 to s33).
[0054]
Search example 4: From the search inheritance relationship table 40 on the condition that “<Name> = Christmas concert AND ID = C”, the child right definition ID {C, D, G, H} of “parent right definition ID = C” Acquire (s11, s14, s17), and obtain a set of records having these in the right definition ID from the search index table 50 (s18). The right definition ID {C, D, G, H} of the record “<Name> = Christmas concert” is acquired from the set of records (*), and this is used as the search result (s19).
[0055]
If TRUE is specified for the child exclusion flag, a search is made for those whose Name flag is TRUE among these results, but since there is no match, {C} is used as the search result (s41 to s47) (* stage (Because the existence of “parent right definition ID = child right definition ID = C” has already been confirmed).
[0056]
A processing procedure for reconfiguring the logical structure of each right definition ID, which is the result obtained in the search example, using the number of inherited stages will be described.
[0057]
For example, in order to acquire the logical structure of right definition ID = H,
(1) A record of “child right definition ID = H” is acquired from the inheritance relationship table 40, and the inheritance relationship is reconfigured by arranging the parent right definition IDs in descending order of the number of inheritance steps.
[0058]
In this case, since {parent right definition ID, number of inheritance stages} = {A, 3} {B, 2} {C, 1} {H, 0}, the inheritance relationship of A-B-C-H in order from the inheritance parent. Is obtained.
[0059]
(2) For each right definition ID, a right structure is obtained from the main table 30 and its contents are connected in order from the top to obtain a logical structure.
[0060]
That is, the right definition with right definition ID = H is identified as the following content information as a logical structure.
[0061]
<Type> Concert </ Type>
<Name> Christmas Concert </ Name>
<Place> Tokyo </ Place>
<Gate> Gate3 </ Gate>
[0062]
【The invention's effect】
As described above, according to the inventions of claims 1 and 3, by using the inheritance relationship table, a reference destination is acquired from the reference information part of the structured document, and the inheritance relationship is not sequentially traced. A set of structured documents inherited by a structured document can be specified. Further, even if the inheritance relationship is not stored in a tree structure in advance, all child structured documents inheriting a certain structured document can be specified by simple table reference.
[0063]
Further, according to the inventions of claims 2 and 3, by managing the number of inheritance stages, the reference information part is sequentially analyzed for a plurality of structured documents inherited by a certain structured document, Without sequentially referencing, it is possible to obtain the inheritance hierarchical structure of all structured documents in which the inheritance relationship of a structured document having a certain document identification information is defined. It is possible to easily reconstruct a document in which the content parts of all structured documents in which inheritance relations are included are connected in the order of inheritance hierarchy.
[0064]
Further, according to the inventions of claims 4 and 5, by using a search index table in which an index for a specific search target area in a structured document is used, the logical structures of all structured documents can be acquired in advance. Also, a partial match search can be performed on the logical structure.
[0065]
Further, according to the inventions of claims 6 and 7, the search index table is provided with a search target area appearance flag, so that the structure having the search target area whose information is lost after the logical structure is reconfigured. The physical entity of the document can be specified.
[Brief description of the drawings]
FIG. 1 is a schematic diagram showing a basic structure of a structured document targeted by the present invention. FIG. 2 is a schematic diagram showing an example of inheritance relationships of structured documents. FIG. Fig. 4 is a schematic diagram showing a specific example of a right definition. Fig. 5 is a schematic diagram of a right definition inheritance tree shown in Fig. 4. Fig. 6 is an example of a device configuration for realizing the method of the present invention. FIG. 7 is a schematic diagram illustrating an example of a main table in FIG. 6. FIG. 8 is a schematic diagram illustrating an example of an inheritance relationship table in FIG. 6. FIG. 9 is an example of a search index table in FIG. FIG. 10 is a flowchart of a document storage program corresponding to the document storage method of claim 1. FIG. 11 is a flowchart of a document storage program corresponding to the document storage method of claim 2. Flow chart of document storage program corresponding to the document storage method [Figure] 3. Flow chart of document storage program corresponding to document storage method of claim 6. [FIG. 14] Flow chart of document storage program corresponding to document search method of claim 3. [FIG. 15] Corresponding to document search method of claim 5. FIG. 16 is a flowchart of a document storage program corresponding to the document search method of claim 7. FIG. 17 is a flowchart of a document storage program corresponding to the document search method of FIG. FIG. 19 is a flowchart of a document storage program corresponding to the document search method of FIG. 7. FIG. 20 is an explanatory diagram illustrating an example of document storage processing.
30: Main table, 31: Document column (right definition column), 32: Document identification information column (right definition ID column), 40: Inheritance relationship table, 41: Parent information column (parent right definition ID column), 42: Child Information column (child rights definition ID column), 43: relative stage number column, 50: search index table, 51: identification information column (right definition ID column), 52: search target area column (Type column, Name column, Place column, Gate column), 53: Appearance flag column (Type flag column, Name flag column, Place flag column, Gate flag column), 60: Document storage processing unit, and 70: Document search processing unit.

Claims

In the database, a method of adding storing a new structured document using computers accessible to the database,
The structured document includes a reference information part and a content part, and the reference information part has zero or no unique document identification information corresponding to each structured document as information for designating a reference destination structured document as an inheritance parent. Contains only one, the content part contains the content information of the structured document,
The database includes a main table including at least one record including a document column storing the structured document and a document identification information column storing document identification information corresponding to the structured document, and a structure serving as an inheritance parent An inheritance relationship table that includes at least one record including a parent information column that stores document identification information corresponding to a structured document and a child information column that stores document identification information corresponding to a structured document serving as an inheritor. And
The computer is
Creating document identification information corresponding to the structured document to be added;
Adding a record having a structured document to be added to the document column to the main table of the database and having document identification information of the structured document to be added to the document identification information column;
Determining whether the reference document identification information is included in the reference information part of the structured document to be added;
If the reference document identification information is not included in the reference information part,
Adding a record having document identification information of a structured document to be added to both the parent information column and the child information column in the inheritance relationship table of the database;
When the reference document identification information is included in the reference information part,
Searching and acquiring a set of records having the document identification information of the reference destination in a child information column from the inheritance relationship table of the database;
In the inheritance relationship table of the database, for each record of the acquired set, the parent information column has the document identification information included in the parent information column of the record, and the child information column has the document identification information of the structured document to be added. Adding a record having
Adding a record having the document identification information of the structured document to be added to both the parent information column and the child information column to the inheritance relationship table of the database.

The structured document storage method according to claim 1,
In addition to the above, each record of the inheritance relationship table stores a relative number of inheritance stages between the structured document corresponding to the document identification information in the parent information column and the structured document corresponding to the document identification information in the child information column. Has a column of number of stages,
Instead of the steps of claim 1,
The computer is
If the reference document identification information is not included in the reference information part,
Adding a record having document identification information of a structured document to be added to both the parent information column and the child information column and having a value of 0 in the relative column number column in the inheritance relationship table of the database ;
When the reference document identification information is included in the reference information part,
Searching and acquiring a set of records having the document identification information of the reference destination in a child information column from the inheritance relationship table of the database ;
In the inheritance relationship table of the database, for each record of the acquired set, the parent information column has the document identification information included in the parent information column of the record, and the child information column has the document identification information of the structured document to be added. And adding a record having a value obtained by adding 1 to the value included in the relative column number column of the record to the relative column number column;
Executing a step of adding a record having document identification information of a structured document to be added to both the parent information column and the child information column and having a value of 0 in the relative column number column to the inheritance relation table of the database. A structured document storage method characterized by:

From the database, a structured document search method for searching a structured document corresponding to the search conditions using computers accessible to the database,
The structured document includes a reference information part and a content part, and the reference information part has zero or no unique document identification information corresponding to each structured document as information for designating a reference destination structured document as an inheritance parent. Contains only one, the content part contains the content information of the structured document,
The database includes a main table including at least one record including a document column storing the structured document and a document identification information column storing document identification information corresponding to the structured document, and a structure serving as an inheritance parent An inheritance relationship table that includes at least one record including a parent information column that stores document identification information corresponding to a structured document and a child information column that stores document identification information corresponding to a structured document serving as an inheritor. And
The computer is
Designating document identification information as the search condition;
Searching and acquiring a set of records having the specified document identification information in a parent information column from the inheritance relationship table of the database;
A structured document search method comprising: performing a search result of document identification information described in a child information column of a record included in the set.

The structured document storage method according to claim 1 or 2,
The database includes a record including an identification information column that stores document identification information and at least one search target area column that stores a value described in the search target area that is a part of content information designated in advance. the stored together search index table containing at least one,
In addition to the steps according to claim 1 or 2,
The computer is
Obtaining a value described in the search target area from the structured document to be added;
If the reference document identification information is not included in the reference information part,
Adding a record having document identification information of a structured document to be added to an identification information column and having a value of each acquired search target area in each search target area column in the search index table of the database ; ,
When the reference document identification information is included in the reference information part,
Searching and obtaining a record having the document identification information of the reference destination in the identification information column from the search index table of the database ;
In the search index table of the database , the identification information column has the document identification information of the structured document to be added, and among each search target area column, the value is stored in the record. And adding a record having the value of the acquired search target area for a column in which no value is stored in the record.

From the database, a structured document search method for searching a structured document corresponding to the search conditions using computers accessible to the database,
The structured document includes a reference information part and a content part, and the reference information part has zero or no unique document identification information corresponding to each structured document as information for designating a reference destination structured document as an inheritance parent. Contains only one, the content part contains the content information of the structured document,
The database includes a main table including at least one record including a document column storing the structured document and a document identification information column storing document identification information corresponding to the structured document, and a structure serving as an inheritance parent An inheritance relationship table including at least one record including a parent information column storing document identification information corresponding to a structured document and a child information column storing document identification information corresponding to a structured document serving as an inheritor; and a document It includes at least one record including an identification information column for storing identification information and at least one search target area column for storing a value described in the search target area that is a part of content information designated in advance. And a search index table
The computer is
As the search condition, document identification information, an AND condition based on a combination of some of the search target areas and a value described in the search target area, or some of the search target areas and the search target area are described. Specifying an AND condition based on a set of values and document identification information;
If only document identification information is specified in the search condition,
Searching and acquiring a set of records having the specified document identification information in a parent information column from the inheritance relationship table of the database;
Making the document identification information described in the child information column of the records included in the set a search result;
When only the AND condition based on a combination of some of the search target areas and the value described in the search target area is specified in the search condition,
Searching and acquiring a set of records having the specified value in the specified search target area column from the search index table of the database;
Making the document identification information described in the identification information column of the records included in the set a search result;
When an AND condition based on a set of some of the search target areas and a value described in the search target area and document identification information is specified in the search condition,
Searching and acquiring a set of records having the specified document identification information in a parent information column from the inheritance relationship table of the database;
From the search index table of the database, the identification information column has document identification information described in the child information of the records included in the set, and the specified search target area column has the specified value. Retrieving and retrieving a set;
A structured document retrieval method comprising: retrieving document identification information described in an identification information column of a record included in the set as a retrieval result.

The structured document storage method according to claim 1 or 2 ,
The database includes an identification information column for storing document identification information, at least one search target area column for storing a value described in a search target area that is a part of content information designated in advance, and each search If the value in the column for each target region column is first written in the contents of the structured document takes a true value, from otherwise the indication flag column for storing the indication flag taking the false value Together with a search index table that contains at least one record
In addition to the steps according to claim 1 or 2,
The computer is
Obtaining a value described in the search target area from the structured document to be added;
If the reference document identification information is not included in the reference information part,
The search index table of the database has document identification information of the structured document to be added in the identification information column, each search target area column has the value of each acquired search target area, and each appearance flag column Adding a record having a true value for a column corresponding to the search target area having the value of the acquired search target area,
When the reference document identification information is included in the reference information part,
Searching and obtaining a record having the document identification information of the reference destination in the identification information column from the search index table of the database ;
In the search index table of the database , the identification information column has the document identification information of the structured document to be added, and among each search target area column, the value is stored in the record. However, the column in which no value is stored in the record has the value of the acquired search target area, and among the appearance flag columns, the column corresponding to the search target area column having the value of the record is false. Adding a record having a true value to a column corresponding to the search target area column having a value and having the value of the acquired search target area.

From the database, a structured document search method for searching a structured document corresponding to the search conditions using computers accessible to the database,
The structured document includes a reference information part and a content part, and the reference information part has zero or no unique document identification information corresponding to each structured document as information for designating a reference destination structured document as an inheritance parent. Contains only one, the content part contains the content information of the structured document,
The database includes a main table including at least one record including a document column storing the structured document and a document identification information column storing document identification information corresponding to the structured document, and a structure serving as an inheritance parent An inheritance relationship table including at least one record including a parent information column storing document identification information corresponding to a structured document and a child information column storing document identification information corresponding to a structured document serving as an inheritor; and a document An identification information column for storing identification information, at least one search target region column for storing a value described in the search target region that is a part of the content information designated in advance, and each search target region column Takes a true value if the value in the column is described for the first time in the content part of the structured document, and takes a false value otherwise. The record consisting of the indication flag column for storing the flag and stores the search index table containing at least one,
The computer is
As the search condition, document identification information, an AND condition based on a combination of some of the search target areas and a value described in the search target area, or some of the search target areas and the search target area are described. Specify the true value when acquiring only the AND condition based on the set of values and the document identification information, and only those at the top of the inheritance relationship that match the above conditions as search targets, otherwise false values Specifying a child exclusion flag that is specified or omitted;
When only document identification information and a child exclusion flag are specified in the search condition and a false value is specified or omitted in the child exclusion flag,
Searching and acquiring a set of records having the specified document identification information in a parent information column from the inheritance relationship table of the database;
Making the document identification information described in the child information column of the records included in the set a search result;
When only document identification information and a child exclusion flag are specified in the search condition and a true value is specified for the child exclusion flag,
Searching for a record having the specified document identification information in both the parent information column and the child information column in the inheritance relationship table of the database; and
Making the document identification information a search result only when a record exists;
When the AND condition and only the child exclusion flag specified by a combination of some of the search target areas and the value described in the search target area are specified in the search condition, and a false value is specified in the child exclusion flag or omitted If
Searching and acquiring a set of records having the specified value in the specified search target area column from the search index table of the database;
Making the document identification information described in the identification information column of the records included in the set a search result;
When the AND condition and only the child exclusion flag specified by a combination of some of the search target areas and the values described in the search target area are specified in the search condition and the true value is specified for the child exclusion flag Is
Searching and acquiring a set of records having the specified value in the specified search target area column from the search index table of the database;
Document identification information described in the identification information column of a record having a column in which at least one of the appearance flag columns corresponding to the specified search target area column is stored among the records included in the set. A search result as a step,
When the AND condition and the child exclusion flag based on the document identification information and a set of some of the search target areas and the values described in the search target area are specified in the search condition, and a false value is specified in the child exclusion flag If omitted or omitted,
Searching and acquiring a set of records having the specified document identification information in a parent information column from the inheritance relationship table of the database;
A record having the document identification information described in the child information column of the record included in the set in the identification information column and the specified value in the specified search target area column from the search index table of the database Retrieving and obtaining a second set of:
Using the document identification information described in the identification information column of the records included in the second set as a search result;
When the AND condition and the child exclusion flag based on the document identification information and a set of some of the search target areas and the values described in the search target area are specified in the search condition, and the true value is specified in the child exclusion flag If
Searching and acquiring a set of records having the specified document identification information in a parent information column from the inheritance relationship table of the database;
From the search index table of the database, the identification information column has document identification information described in the child information column of the record included in the set, and the specified search target area column has the specified value. Searching and acquiring a second set of records having columns in which at least one of the appearance flag columns corresponding to the designated search target area column is stored;
If there is an element in the second set, the document identification information described in the identification information column of the record is taken as a search result;
If there is no element in the second set, searching for a record having the specified document identification information in both the parent information column and the child information column in the inheritance relationship table of the database;
A structured document search method comprising: a step of using the document identification information as a search result only when a record exists.